01_neural_network_regression_with_tensorflow_video.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/fyemane/TensorFlow-ZTM-/blob/main/01_neural_network_regression_with_tensorflow_video.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wJ-DtfSBAO_4"
      },
      "source": [
        "# Introduction to Regression with Neural Networks in TensorFlow\n",
        "\n",
        "There are many definitions for a regression problem but in our case, we're going to simplify it: predicting a numerical variable based on some other combination of variables, even shorter... predicting a number."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gb19CXPVAv-R",
        "outputId": "78611a2c-9ab7-4339-a459-fea411d0c034"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.15.0\n"
          ]
        }
      ],
      "source": [
        "# Import tensorflow\n",
        "import tensorflow as tf\n",
        "print(tf.__version__)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aGyXEYakA0r6"
      },
      "source": [
        "## Creating sample regression data (so we can model it)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 435
        },
        "id": "M8PiP6tLA_Vt",
        "outputId": "1bfa246c-b827-46ec-ef7e-cac63c485c0c"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAicAAAGiCAYAAAA8xWYrAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAd7klEQVR4nO3df2zU93348dfZFDvtzGUmmDs3hhrakrqUbHQ1Q0ujRSHBTPJC20lNVKYwRdmGSLaEdl0zJXW8VaPJpCjqlBFt0hpFLOk2aaWi0yx1ZICi8kMLQ5XFGgXkKESxYQNxBibT1P58/0jxF2Pzw3D43j4/HtJJuc/n47tXdDr5yX3u83Yuy7IsAAASUVPpAQAALiROAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKRMKk42b94cn/vc56KhoSGamppi7dq18eabb4455jd/8zcjl8uNuf3hH/5hWYcGAKrXpOJk165dsXHjxti7d2/86Ec/ivfffz/uvffeOHv27JjjHn744ejv7x+9Pfvss2UdGgCoXrMmc3BPT8+Y+y+99FI0NTXFG2+8EXfeeefo9g9/+MNRKBTKMyEAMKNMKk4uViqVIiKisbFxzPZ/+Id/iK1bt0ahUIjOzs546qmn4sMf/vCEj3Hu3Lk4d+7c6P2RkZE4efJkzJ07N3K53PWMBwBMkSzL4vTp09Hc3Bw1Ndf3ldZclmXZtfzgyMhI/PZv/3acOnUqXn/99dHtf/u3fxsLFy6M5ubm+MlPfhJ/+qd/Gu3t7fEv//IvEz7O008/Hd3d3dc2PQCQlKNHj8att956XY9xzXGyYcOG+Ld/+7d4/fXXLzvEa6+9FnfffXccPnw4Fi9ePG7/xZ+clEqlWLBgQRw9ejTmzJlzLaMBAFNscHAwWlpa4tSpU5HP56/rsa7ptM4jjzwSP/zhD2P37t1XrKMVK1ZERFwyTurq6qKurm7c9jlz5ogTAJhmyvGVjEnFSZZl8eijj8b3v//92LlzZ7S2tl7xZw4ePBgREcVi8ZoGBABmlknFycaNG+OVV16JH/zgB9HQ0BADAwMREZHP5+Omm26KI0eOxCuvvBK/9Vu/FXPnzo2f/OQn8fjjj8edd94Zy5YtuyH/AwBAdZnUd04u9VHNd7/73Vi/fn0cPXo01q1bF729vXH27NloaWmJL3zhC/Hkk09e9SmawcHByOfzUSqVnNYBgGminL+/J31a53JaWlpi165d1zUQADCz+ds6AEBSxAkAkBRxAgAkRZwAAEm5rr+tAwBMH8MjWezvOxnHTw9FU0N9tLc2Rm1Nen/HTpwAwAzQ09sf3dsPRX9paHRbMV8fXZ1t0bE0rYVSndYBgCrX09sfG7YeGBMmEREDpaHYsPVA9PT2V2iyiYkTAKhiwyNZdG8/FBOtVHZ+W/f2QzE8ck1/B/iGECcAUMX2950c94nJhbKI6C8Nxf6+k1M31BWIEwCoYsdPXzpMruW4qSBOAKCKNTXUl/W4qSBOAKCKtbc2RjFfH5e6YDgXH1y1097aOJVjXZY4AYAqVluTi67OtoiIcYFy/n5XZ1tS652IEwCoch1Li7Fl3fIo5Meeuink62PLuuXJrXNiETYAmAE6lhbjnraCFWIBgHTU1uRi5eK5lR7jipzWAQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASMqsSg8AAFNheCSL/X0n4/jpoWhqqI/21saorclVeiwmIE4AqHo9vf3Rvf1Q9JeGRrcV8/XR1dkWHUuLFZyMiTitA0BV6+ntjw1bD4wJk4iIgdJQbNh6IHp6+ys0GZciTgCoWsMjWXRvPxTZBPvOb+vefiiGRyY6gkoRJwBUrf19J8d9YnKhLCL6S0Oxv+/k1A3FFYkTAKrW8dOXDpNrOY6pIU4AqFpNDfVlPY6pIU4AqFrtrY1RzNfHpS4YzsUHV+20tzZO5VhcgTgBoGrV1uSiq7MtImJcoJy/39XZZr2TxIgTAKpax9JibFm3PAr5saduCvn62LJuuXVOEmQRNgCqXsfSYtzTVrBC7DQhTgCYEWprcrFy8dxKj8FVcFoHAEiKOAEAkiJOAICkiBMAICniBABIyqTiZPPmzfG5z30uGhoaoqmpKdauXRtvvvnmmGOGhoZi48aNMXfu3PilX/ql+NKXvhTHjh0r69AAQPWaVJzs2rUrNm7cGHv37o0f/ehH8f7778e9994bZ8+eHT3m8ccfj+3bt8c///M/x65du+K9996LL37xi2UfHACoTrksy7Jr/eH/+Z//iaampti1a1fceeedUSqVYt68efHKK6/E7/zO70RExE9/+tP41Kc+FXv27Ilf//Vfv+JjDg4ORj6fj1KpFHPmzLnW0QCAKVTO39/X9Z2TUqkUERGNjR/8waQ33ngj3n///Vi1atXoMbfddlssWLAg9uzZM+FjnDt3LgYHB8fcAICZ65rjZGRkJB577LH4jd/4jVi6dGlERAwMDMTs2bPj5ptvHnPs/PnzY2BgYMLH2bx5c+Tz+dFbS0vLtY4EAFSBa46TjRs3Rm9vb3zve9+7rgGeeOKJKJVKo7ejR49e1+MBANPbNf1tnUceeSR++MMfxu7du+PWW28d3V4oFOJnP/tZnDp1asynJ8eOHYtCoTDhY9XV1UVdXd21jAEAVKFJfXKSZVk88sgj8f3vfz9ee+21aG1tHbP/s5/9bHzoQx+KHTt2jG57880345133omVK1eWZ2IAoKpN6pOTjRs3xiuvvBI/+MEPoqGhYfR7JPl8Pm666abI5/Px0EMPxaZNm6KxsTHmzJkTjz76aKxcufKqrtQBAJjUpcS5XG7C7d/97ndj/fr1EfHBImxf/epX49VXX41z587F6tWr42/+5m8ueVrnYi4lBoDpp5y/v69rnZMbQZwAwPSTzDonAADlJk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIyqxKDwDA1BgeyWJ/38k4fnoomhrqo721MWprcpUeC8YRJwAzQE9vf3RvPxT9paHRbcV8fXR1tkXH0mIFJ4PxnNYBqHI9vf2xYeuBMWESETFQGooNWw9ET29/hSaDiYkTgCo2PJJF9/ZDkU2w7/y27u2HYnhkoiOgMsQJQBXb33dy3CcmF8oior80FPv7Tk7dUHAF4gSgih0/fekwuZbjYCqIE4Aq1tRQX9bjYCqIE4Aq1t7aGMV8fVzqguFcfHDVTntr41SOBZclTgCqWG1NLro62yIixgXK+ftdnW3WOyEp4gSgynUsLcaWdcujkB976qaQr48t65Zb54TkWIQNYAboWFqMe9oKVohlWhAnADNEbU0uVi6eW+kx4Iqc1gEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEjKrEoPADBVhkey2N93Mo6fHoqmhvpob22M2ppcpccCLjLpT052794dnZ2d0dzcHLlcLrZt2zZm//r16yOXy425dXR0lGtegGvS09sfdzzzWjzwd3vjj793MB74u71xxzOvRU9vf6VHAy4y6Tg5e/Zs3H777fHCCy9c8piOjo7o7+8fvb366qvXNSTA9ejp7Y8NWw9Ef2lozPaB0lBs2HpAoEBiJn1aZ82aNbFmzZrLHlNXVxeFQuGahwIol+GRLLq3H4psgn1ZROQionv7obinreAUDyTihnwhdufOndHU1BRLliyJDRs2xIkTJy557Llz52JwcHDMDaBc9vedHPeJyYWyiOgvDcX+vpNTNxRwWWWPk46Ojnj55Zdjx44d8cwzz8SuXbtizZo1MTw8POHxmzdvjnw+P3praWkp90jADHb89KXD5FqOA268sl+tc//994/+92c+85lYtmxZLF68OHbu3Bl33333uOOfeOKJ2LRp0+j9wcFBgQKUTVNDfVmPA268G77OyaJFi+KWW26Jw4cPT7i/rq4u5syZM+YGUC7trY1RzNfHpb5NkouIYv6Dy4qBNNzwOHn33XfjxIkTUSwWb/RTAYxTW5OLrs62iIhxgXL+fldnmy/DQkImHSdnzpyJgwcPxsGDByMioq+vLw4ePBjvvPNOnDlzJv7kT/4k9u7dG2+//Xbs2LEj7rvvvvj4xz8eq1evLvfsAFelY2kxtqxbHoX82FM3hXx9bFm3PDqW+scTpCSXZdlEV9hd0s6dO+Ouu+4at/3BBx+MLVu2xNq1a+O//uu/4tSpU9Hc3Bz33ntv/MVf/EXMnz//qh5/cHAw8vl8lEolp3iAsrJCLNw45fz9Pek4udHECQBMP+X8/e0P/wEASREnAEBSxAkAkBRxAgAkRZwAAEkRJwBAUsQJAJAUcQIAJEWcAABJEScAQFLECQCQFHECACRFnAAASREnAEBSxAkAkBRxAgAkRZwAAEkRJwBAUsQJAJAUcQIAJEWcAABJEScAQFLECQCQFHECACRFnAAASREnAEBSxAkAkBRxAgAkRZwAAEkRJwBAUsQJAJAUcQIAJEWcAABJEScAQFLECQCQFHECACRlVqUHAKbG8EgW+/tOxvHTQ9HUUB/trY1RW5Or9FgA44gTmAF6evuje/uh6C8NjW4r5uujq7MtOpYWKzgZwHhO60CV6+ntjw1bD4wJk4iIgdJQbNh6IHp6+ys0GcDExAlUseGRLLq3H4psgn3nt3VvPxTDIxMdAVAZ4gSq2P6+k+M+MblQFhH9paHY33dy6oYCuAJxAlXs+OlLh8m1HAcwFcQJVLGmhvqyHgcwFcQJVLH21sYo5uvjUhcM5+KDq3baWxunciyAyxInUMVqa3LR1dkWETEuUM7f7+pss94JkBRxAlWuY2kxtqxbHoX82FM3hXx9bFm33DonQHIswgYzQMfSYtzTVrBCLDAtiBOYIWprcrFy8dxKjwFwRU7rAABJEScAQFLECQCQFHECACRFnAAASREnAEBSxAkAkBRxAgAkRZwAAEkRJwBAUsQJAJAUcQIAJEWcAABJEScAQFLECQCQFHECACRFnAAASREnAEBSxAkAkBRxAgAkRZwAAEmZdJzs3r07Ojs7o7m5OXK5XGzbtm3M/izL4pvf/GYUi8W46aabYtWqVfHWW2+Va14AoMpNOk7Onj0bt99+e7zwwgsT7n/22WfjO9/5Trz44ouxb9+++MhHPhKrV6+OoaGh6x4WAKh+syb7A2vWrIk1a9ZMuC/Lsnj++efjySefjPvuuy8iIl5++eWYP39+bNu2Le6///7rmxYAqHpl/c5JX19fDAwMxKpVq0a35fP5WLFiRezZs2fCnzl37lwMDg6OuQEAM1dZ42RgYCAiIubPnz9m+/z580f3XWzz5s2Rz+dHby0tLeUcCQCYZip+tc4TTzwRpVJp9Hb06NFKjwQAVFBZ46RQKERExLFjx8ZsP3bs2Oi+i9XV1cWcOXPG3ACAmauscdLa2hqFQiF27Ngxum1wcDD27dsXK1euLOdTAQBVatJX65w5cyYOHz48er+vry8OHjwYjY2NsWDBgnjsscfiW9/6VnziE5+I1tbWeOqpp6K5uTnWrl1bzrkBgCo16Tj5z//8z7jrrrtG72/atCkiIh588MF46aWX4utf/3qcPXs2fv/3fz9OnToVd9xxR/T09ER9fX35pgYAqlYuy7Ks0kNcaHBwMPL5fJRKJd8/AYBpopy/vyt+tQ4AwIXECQCQFHECACRFnAAASZn01TowXQ2PZLG/72QcPz0UTQ310d7aGLU1uUqPBcBFxAkzQk9vf3RvPxT9paHRbcV8fXR1tkXH0mIFJwPgYk7rUPV6evtjw9YDY8IkImKgNBQbth6Int7+Ck0GwETECVVteCSL7u2HYqLFfM5v695+KIZHklruB2BGEydUtf19J8d9YnKhLCL6S0Oxv+/k1A0FwGWJE6ra8dOXDpNrOQ6AG0+cUNWaGq7ubzpd7XEA3HjihKrW3toYxXx9XOqC4Vx8cNVOe2vjVI4FwGWIE6pabU0uujrbIiLGBcr5+12dbdY7AUiIOKHqdSwtxpZ1y6OQH3vqppCvjy3rllvnBCAxFmFjRuhYWox72gpWiAWYBsQJM0ZtTS5WLp5b6TEAuAKndQCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIyq9IDMDWGR7LY33cyjp8eiqaG+mhvbYzamlylxwKAccTJDNDT2x/d2w9Ff2lodFsxXx9dnW3RsbRYwckAYDyndapcT29/bNh6YEyYREQMlIZiw9YD0dPbX6HJAGBi4qSKDY9k0b39UGQT7Du/rXv7oRgemegIAKgMcVLF9vedHPeJyYWyiOgvDcX+vpNTNxQAXIE4qWLHT186TK7lOACYCuKkijU11Jf1OACYCuKkirW3NkYxXx+XumA4Fx9ctdPe2jiVYwHAZYmTKlZbk4uuzraIiHGBcv5+V2eb9U4ASIo4qXIdS4uxZd3yKOTHnrop5Otjy7rl1jkBIDkWYZsBOpYW4562ghViAZgWxMkMUVuTi5WL51Z6DAC4Iqd1AICkiBMAICniBABIijgBAJIiTgCApJQ9Tp5++unI5XJjbrfddlu5nwYAqFI35FLiT3/60/Hv//7v//9JZrliGQC4OjekGmbNmhWFQuFGPDQAUOVuyHdO3nrrrWhubo5FixbFV77ylXjnnXcueey5c+dicHBwzA0AmLnKHicrVqyIl156KXp6emLLli3R19cXn//85+P06dMTHr958+bI5/Ojt5aWlnKPBABMI7ksy7Ib+QSnTp2KhQsXxnPPPRcPPfTQuP3nzp2Lc+fOjd4fHByMlpaWKJVKMWfOnBs5GgBQJoODg5HP58vy+/uGf1P15ptvjk9+8pNx+PDhCffX1dVFXV3djR4DAJgmbvg6J2fOnIkjR45EsVi80U8FAFSBssfJ1772tdi1a1e8/fbb8eMf/zi+8IUvRG1tbTzwwAPlfioAoAqV/bTOu+++Gw888ECcOHEi5s2bF3fccUfs3bs35s2bV+6nAgCqUNnj5Hvf+165HxIAmEH8bR0AICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKSIEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKTMqvQAU2V4JIv9fSfj+OmhaGqoj/bWxqityVV6LADgIjMiTnp6+6N7+6HoLw2Nbivm66Orsy06lhYrOBkAcLGqP63T09sfG7YeGBMmEREDpaHYsPVA9PT2V2gyAGAiVR0nwyNZdG8/FNkE+85v695+KIZHJjoCAKiEqo6T/X0nx31icqEsIvpLQ7G/7+TUDQUAXFZVx8nx05cOk2s5DgC48ao6Tpoa6st6HABw41V1nLS3NkYxXx+XumA4Fx9ctdPe2jiVYwEAl1HVcVJbk4uuzraIiHGBcv5+V2eb9U4AICFVHScRER1Li7Fl3fIo5Meeuink62PLuuXWOQGAxMyIRdg6lhbjnraCFWIBYBqYEXES8cEpnpWL51Z6DADgCqr+tA4AML2IEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApCS3QmyWZRERMTg4WOFJAICrdf739vnf49cjuTg5ffp0RES0tLRUeBIAYLJOnz4d+Xz+uh4jl5UjccpoZGQk3nvvvWhoaIhcbub+Yb7BwcFoaWmJo0ePxpw5cyo9DpfhtZpevF7Th9dq+jj/Wh06dCiWLFkSNTXX962R5D45qampiVtvvbXSYyRjzpw53pTThNdqevF6TR9eq+njox/96HWHSYQvxAIAiREnAEBSxEmi6urqoqurK+rq6io9ClfgtZpevF7Th9dq+ij3a5XcF2IBgJnNJycAQFLECQCQFHECACRFnAAASREn08DHPvaxyOVyY27f/va3Kz0Wv/DCCy/Exz72saivr48VK1bE/v37Kz0SF3n66afHvYduu+22So/FL+zevTs6Ozujubk5crlcbNu2bcz+LMvim9/8ZhSLxbjpppti1apV8dZbb1Vm2BnuSq/V+vXrx73XOjo6Jv084mSa+PM///Po7+8fvT366KOVHomI+Md//MfYtGlTdHV1xYEDB+L222+P1atXx/Hjxys9Ghf59Kc/PeY99Prrr1d6JH7h7Nmzcfvtt8cLL7ww4f5nn302vvOd78SLL74Y+/bti4985COxevXqGBoamuJJudJrFRHR0dEx5r326quvTvp5klu+nok1NDREoVCo9Bhc5LnnnouHH344fu/3fi8iIl588cX413/91/j7v//7+MY3vlHh6bjQrFmzvIcStWbNmlizZs2E+7Isi+effz6efPLJuO+++yIi4uWXX4758+fHtm3b4v7775/KUWe8y71W59XV1V33e80nJ9PEt7/97Zg7d2786q/+avzVX/1V/PznP6/0SDPez372s3jjjTdi1apVo9tqampi1apVsWfPngpOxkTeeuutaG5ujkWLFsVXvvKVeOeddyo9Elehr68vBgYGxrzP8vl8rFixwvssUTt37oympqZYsmRJbNiwIU6cODHpx/DJyTTwR3/0R7F8+fJobGyMH//4x/HEE09Ef39/PPfcc5UebUb73//93xgeHo758+eP2T5//vz46U9/WqGpmMiKFSvipZdeiiVLlkR/f390d3fH5z//+ejt7Y2GhoZKj8dlDAwMRERM+D47v490dHR0xBe/+MVobW2NI0eOxJ/92Z/FmjVrYs+ePVFbW3vVjyNOKuQb3/hGPPPMM5c95r//+7/jtttui02bNo1uW7ZsWcyePTv+4A/+IDZv3mxZZ7gKF34MvWzZslixYkUsXLgw/umf/ikeeuihCk4G1eXC02yf+cxnYtmyZbF48eLYuXNn3H333Vf9OOKkQr761a/G+vXrL3vMokWLJty+YsWK+PnPfx5vv/12LFmy5AZMx9W45ZZbora2No4dOzZm+7Fjx3y3IXE333xzfPKTn4zDhw9XehSu4Px76dixY1EsFke3Hzt2LH7lV36lQlNxtRYtWhS33HJLHD58WJxMB/PmzYt58+Zd088ePHgwampqoqmpqcxTMRmzZ8+Oz372s7Fjx45Yu3ZtRESMjIzEjh074pFHHqnscFzWmTNn4siRI/G7v/u7lR6FK2htbY1CoRA7duwYjZHBwcHYt29fbNiwobLDcUXvvvtunDhxYkxYXg1xkrg9e/bEvn374q677oqGhobYs2dPPP7447Fu3br45V/+5UqPN+Nt2rQpHnzwwfi1X/u1aG9vj+effz7Onj07evUOafja174WnZ2dsXDhwnjvvfeiq6sramtr44EHHqj0aMQHsXjhp1h9fX1x8ODBaGxsjAULFsRjjz0W3/rWt+ITn/hEtLa2xlNPPRXNzc2j/yhg6lzutWpsbIzu7u740pe+FIVCIY4cORJf//rX4+Mf/3isXr16ck+UkbQ33ngjW7FiRZbP57P6+vrsU5/6VPaXf/mX2dDQUKVH4xf++q//OluwYEE2e/bsrL29Pdu7d2+lR+IiX/7yl7NisZjNnj07++hHP5p9+ctfzg4fPlzpsfiF//iP/8giYtztwQcfzLIsy0ZGRrKnnnoqmz9/flZXV5fdfffd2ZtvvlnZoWeoy71W//d//5fde++92bx587IPfehD2cKFC7OHH344GxgYmPTz5LIsy8qSUwAAZWCdEwAgKeIEAEiKOAEAkiJOAICkiBMAICniBABIijgBAJIiTgCApIgTACAp4gQASIo4AQCSIk4AgKT8P0zcr4hLXzkpAAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Create features\n",
        "X = np.array([-7.0, -4.0, -1.0, 2.0, 5.0, 8.0, 11.0, 14.0])\n",
        "\n",
        "# Create labels\n",
        "y = np.array([3.0, 6.0, 9.0, 12.0, 15.0, 18.0, 21.0, 24.0])\n",
        "\n",
        "# Visualize it\n",
        "plt.scatter(X, y);"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bYtodNxhBZJb",
        "outputId": "42bdd019-2728-4aea-83cc-26b24d506571"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ True,  True,  True,  True,  True,  True,  True,  True])"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ],
      "source": [
        "y == X + 10"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NSA3cD7WBoqU"
      },
      "source": [
        "### Input and output shapes"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WsQiJiRnBz9x",
        "outputId": "55c80524-8cd3-47f4-afdf-916e6dd93295"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(3,), dtype=string, numpy=array([b'bedroom', b'bathroom', b'garage'], dtype=object)>,\n",
              " <tf.Tensor: shape=(1,), dtype=int32, numpy=array([959700], dtype=int32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "# Create a demo tensor for our housing price prediction problem\n",
        "house_info = tf.constant([\"bedroom\", \"bathroom\", \"garage\"])\n",
        "house_price = tf.constant([959700])\n",
        "house_info, house_price"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c7Sv3LrXCh6c",
        "outputId": "eb921cda-8662-49f5-8693-22401a37906f"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((-7.0, 3.0), (-4.0, 6.0))"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ],
      "source": [
        "(X[0], y[0]), (X[1], y[1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vw5AIXOcCNho",
        "outputId": "cdfb1fac-d180-4ad1-e577-a8a4add35f3d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((), ())"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "input_shape = X[0].shape\n",
        "output_shape = y[0].shape\n",
        "input_shape, output_shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QO4yaa4oCcwK",
        "outputId": "bc4037b5-5056-4103-f021-ae7523e7fb84"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "X[0].ndim"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FLIWWvyRB4ua",
        "outputId": "3d63eaed-7b60-4f39-a3bb-c0db5f418e52"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(8,), dtype=float32, numpy=array([-7., -4., -1.,  2.,  5.,  8., 11., 14.], dtype=float32)>,\n",
              " <tf.Tensor: shape=(8,), dtype=float32, numpy=array([ 3.,  6.,  9., 12., 15., 18., 21., 24.], dtype=float32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ],
      "source": [
        "# Turn our NumPy arrays into tensors with dtype float32\n",
        "X = tf.cast(tf.constant(X), dtype=tf.float32)\n",
        "y = tf.cast(tf.constant(y), dtype=tf.float32)\n",
        "X, y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EaPbfwa6B7NM",
        "outputId": "4fd8de4c-3783-4cdf-fe51-0c0548625d94"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(TensorShape([]), TensorShape([]))"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ],
      "source": [
        "input_shape = X[0].shape\n",
        "output_shape = y[0].shape\n",
        "input_shape, output_shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 435
        },
        "id": "8Qjgpel1CDEg",
        "outputId": "a4c0996c-3ccc-40d5-e265-9b3977be1ac9"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.scatter(X, y);"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v2J8QUe4CJZC"
      },
      "source": [
        "## The major steps in modelling with TensorFlow\n",
        "\n",
        "1. Get data ready (turn into tensors)\n",
        "2. Build or pick a pretrained model (to suit your problem)\n",
        "  * Creating a model - define the input and output layers, as well as the hidden layers of a deep learning model.\n",
        "  * Compiling a model - define the loss function (in other words, the function which tells our model how wrong it is), the optimizer (tells our model how to improve the patterns its learning), and evaluation metrics (what we can use to interpret the performance of our model)\n",
        "3. Fit the model to the data and make a prediction\n",
        "  * Fitting a model - letting the model try to find patterns between X & y (features and labels).\n",
        "4. Evaluate the model\n",
        "5. Improve through experimentation\n",
        "6. Save and reload your trained model\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KqLWBggkCYJN",
        "outputId": "014bf78c-51f0-4d48-81c8-db0f0b0c63f2"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "1/1 [==============================] - 0s 466ms/step - loss: 13.6126 - mae: 13.6126\n",
            "Epoch 2/5\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 13.4801 - mae: 13.4801\n",
            "Epoch 3/5\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.3476 - mae: 13.3476\n",
            "Epoch 4/5\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.2151 - mae: 13.2151\n",
            "Epoch 5/5\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.0826 - mae: 13.0826\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba288eb36a0>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ],
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# Create a model using the Sequential API\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/Sequential\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# Compile the model\n",
        "model.compile(loss=tf.keras.losses.mae, # mae is short for mean absolute error\n",
        "              # https://www.tensorflow.org/api_docs/python/tf/keras/losses/MeanAbsoluteError\n",
        "              optimizer=tf.keras.optimizers.SGD(), # SGU is short for stochastic gradient descent\n",
        "              # https://www.tensorflow.org/api_docs/python/tf/keras/optimizers/experimental/SGD\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# Fit the model\n",
        "model.fit(tf.expand_dims(X, axis=-1), y, epochs=5) # epochs = # of times model traines on data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XvJNWMntEfSv",
        "outputId": "4b20e04b-f0f7-47ca-a8ab-020b23573d84"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(8,), dtype=float32, numpy=array([-7., -4., -1.,  2.,  5.,  8., 11., 14.], dtype=float32)>,\n",
              " <tf.Tensor: shape=(8,), dtype=float32, numpy=array([ 3.,  6.,  9., 12., 15., 18., 21., 24.], dtype=float32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ],
      "source": [
        "# Check out X and y\n",
        "X, y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UnFoWwwNJTMS",
        "outputId": "070ac02f-4ca2-4e67-9499-97147e5f3352"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 98ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[2.4780219]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ],
      "source": [
        "# Try and make a prediction using our model\n",
        "y_pred = model.predict(([17.0]))\n",
        "y_pred"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HRkw8aIFJW74",
        "outputId": "49d4737b-97f8-4f49-8388-939b5f2b679a"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[13.478022]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ],
      "source": [
        "y_pred + 11"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "44CoxfEhJpB-"
      },
      "source": [
        "## Steps in improving a model with TensorFlow\n",
        "\n",
        "We can improve our model, by altering the steps we took to create a model.\n",
        "\n",
        "1. **Creating a model** - Here we might add more layers, increase the number of hidden units (also called neurons) within each of the hidden layers, change the activation functions of each layer.\n",
        "2. **Compiling a model** - Here we might change the optimization function or perhaps the **learning rate** of the optimization function.\n",
        "3. **Fitting a model** - Here we might fit a model for more **epochs** (leave it training for longer) or on more data (give the model more examples to learn from).\n",
        "\n",
        "Learning rate: When Adam optimizer tells function to improve, how much should it improve at each step."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X6gfsEiqFwes",
        "outputId": "a7ed3077-6b64-4317-a4c2-017543f63418"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "1/1 [==============================] - 0s 345ms/step - loss: 19.7146 - mae: 19.7146\n",
            "Epoch 2/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 19.4333 - mae: 19.4333\n",
            "Epoch 3/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 19.1521 - mae: 19.1521\n",
            "Epoch 4/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 18.8708 - mae: 18.8708\n",
            "Epoch 5/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 18.5896 - mae: 18.5896\n",
            "Epoch 6/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 18.3083 - mae: 18.3083\n",
            "Epoch 7/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 18.0271 - mae: 18.0271\n",
            "Epoch 8/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 17.7458 - mae: 17.7458\n",
            "Epoch 9/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 17.4646 - mae: 17.4646\n",
            "Epoch 10/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 17.1833 - mae: 17.1833\n",
            "Epoch 11/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 16.9021 - mae: 16.9021\n",
            "Epoch 12/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 16.6208 - mae: 16.6208\n",
            "Epoch 13/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 16.3396 - mae: 16.3396\n",
            "Epoch 14/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 16.0583 - mae: 16.0583\n",
            "Epoch 15/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 15.7771 - mae: 15.7771\n",
            "Epoch 16/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 15.4958 - mae: 15.4958\n",
            "Epoch 17/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 15.2146 - mae: 15.2146\n",
            "Epoch 18/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 14.9333 - mae: 14.9333\n",
            "Epoch 19/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 14.7005 - mae: 14.7005\n",
            "Epoch 20/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 14.5680 - mae: 14.5680\n",
            "Epoch 21/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 14.4355 - mae: 14.4355\n",
            "Epoch 22/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 14.3030 - mae: 14.3030\n",
            "Epoch 23/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 14.1705 - mae: 14.1705\n",
            "Epoch 24/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 14.0380 - mae: 14.0380\n",
            "Epoch 25/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 13.9055 - mae: 13.9055\n",
            "Epoch 26/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.7730 - mae: 13.7730\n",
            "Epoch 27/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.6405 - mae: 13.6405\n",
            "Epoch 28/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.5080 - mae: 13.5080\n",
            "Epoch 29/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.3755 - mae: 13.3755\n",
            "Epoch 30/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 13.2430 - mae: 13.2430\n",
            "Epoch 31/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 13.1105 - mae: 13.1105\n",
            "Epoch 32/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 12.9780 - mae: 12.9780\n",
            "Epoch 33/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 12.8455 - mae: 12.8455\n",
            "Epoch 34/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 12.7130 - mae: 12.7130\n",
            "Epoch 35/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 12.5805 - mae: 12.5805\n",
            "Epoch 36/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 12.4480 - mae: 12.4480\n",
            "Epoch 37/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 12.3155 - mae: 12.3155\n",
            "Epoch 38/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 12.1830 - mae: 12.1830\n",
            "Epoch 39/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 12.0505 - mae: 12.0505\n",
            "Epoch 40/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 11.9180 - mae: 11.9180\n",
            "Epoch 41/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 11.7855 - mae: 11.7855\n",
            "Epoch 42/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 11.6530 - mae: 11.6530\n",
            "Epoch 43/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 11.5205 - mae: 11.5205\n",
            "Epoch 44/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 11.3880 - mae: 11.3880\n",
            "Epoch 45/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 11.2555 - mae: 11.2555\n",
            "Epoch 46/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 11.1230 - mae: 11.1230\n",
            "Epoch 47/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 10.9905 - mae: 10.9905\n",
            "Epoch 48/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 10.8580 - mae: 10.8580\n",
            "Epoch 49/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 10.7255 - mae: 10.7255\n",
            "Epoch 50/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 10.5930 - mae: 10.5930\n",
            "Epoch 51/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 10.4605 - mae: 10.4605\n",
            "Epoch 52/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 10.3280 - mae: 10.3280\n",
            "Epoch 53/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 10.1955 - mae: 10.1955\n",
            "Epoch 54/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 10.0630 - mae: 10.0630\n",
            "Epoch 55/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.9305 - mae: 9.9305\n",
            "Epoch 56/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.7980 - mae: 9.7980\n",
            "Epoch 57/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 9.6655 - mae: 9.6655\n",
            "Epoch 58/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.5330 - mae: 9.5330\n",
            "Epoch 59/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.4005 - mae: 9.4005\n",
            "Epoch 60/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 9.2680 - mae: 9.2680\n",
            "Epoch 61/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.1355 - mae: 9.1355\n",
            "Epoch 62/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 9.0030 - mae: 9.0030\n",
            "Epoch 63/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 8.8705 - mae: 8.8705\n",
            "Epoch 64/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 8.7380 - mae: 8.7380\n",
            "Epoch 65/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 8.6055 - mae: 8.6055\n",
            "Epoch 66/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 8.4730 - mae: 8.4730\n",
            "Epoch 67/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 8.3405 - mae: 8.3405\n",
            "Epoch 68/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 8.2080 - mae: 8.2080\n",
            "Epoch 69/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 8.0755 - mae: 8.0755\n",
            "Epoch 70/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.9430 - mae: 7.9430\n",
            "Epoch 71/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 7.8105 - mae: 7.8105\n",
            "Epoch 72/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 7.6780 - mae: 7.6780\n",
            "Epoch 73/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 7.5455 - mae: 7.5455\n",
            "Epoch 74/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 7.4130 - mae: 7.4130\n",
            "Epoch 75/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.2805 - mae: 7.2805\n",
            "Epoch 76/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 7.1480 - mae: 7.1480\n",
            "Epoch 77/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.0155 - mae: 7.0155\n",
            "Epoch 78/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9563 - mae: 6.9563\n",
            "Epoch 79/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9506 - mae: 6.9506\n",
            "Epoch 80/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9450 - mae: 6.9450\n",
            "Epoch 81/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9394 - mae: 6.9394\n",
            "Epoch 82/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9338 - mae: 6.9338\n",
            "Epoch 83/100\n",
            "1/1 [==============================] - 0s 14ms/step - loss: 6.9281 - mae: 6.9281\n",
            "Epoch 84/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9225 - mae: 6.9225\n",
            "Epoch 85/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9169 - mae: 6.9169\n",
            "Epoch 86/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.9113 - mae: 6.9113\n",
            "Epoch 87/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9056 - mae: 6.9056\n",
            "Epoch 88/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.9000 - mae: 6.9000\n",
            "Epoch 89/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8944 - mae: 6.8944\n",
            "Epoch 90/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8888 - mae: 6.8888\n",
            "Epoch 91/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8831 - mae: 6.8831\n",
            "Epoch 92/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 6.8775 - mae: 6.8775\n",
            "Epoch 93/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 6.8719 - mae: 6.8719\n",
            "Epoch 94/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 6.8663 - mae: 6.8663\n",
            "Epoch 95/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8606 - mae: 6.8606\n",
            "Epoch 96/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8550 - mae: 6.8550\n",
            "Epoch 97/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8494 - mae: 6.8494\n",
            "Epoch 98/100\n",
            "1/1 [==============================] - 0s 8ms/step - loss: 6.8438 - mae: 6.8438\n",
            "Epoch 99/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8381 - mae: 6.8381\n",
            "Epoch 100/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.8325 - mae: 6.8325\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba28855e6e0>"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ],
      "source": [
        "# Let's rebuild our model\n",
        "\n",
        "# 1. Create the model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model.fit(tf.expand_dims(X, axis=-1), y, epochs=100)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "a83XkhvkLvnI",
        "outputId": "2337317a-6a36-4801-dca9-8aa718d1bde3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 83ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[29.515587]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ],
      "source": [
        "# Let's see if our model's prediction has improved...\n",
        "model.predict([17.0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O2ksDygCM4d0",
        "outputId": "634a8a14-73ca-4ab9-c3d5-bf23e4730ed0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "1/1 [==============================] - 1s 848ms/step - loss: 13.0727 - mae: 13.0727\n",
            "Epoch 2/100\n",
            "1/1 [==============================] - 0s 19ms/step - loss: 12.3797 - mae: 12.3797\n",
            "Epoch 3/100\n",
            "1/1 [==============================] - 0s 16ms/step - loss: 11.6758 - mae: 11.6758\n",
            "Epoch 4/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 10.9595 - mae: 10.9595\n",
            "Epoch 5/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 10.2301 - mae: 10.2301\n",
            "Epoch 6/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 9.4861 - mae: 9.4861\n",
            "Epoch 7/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 8.7257 - mae: 8.7257\n",
            "Epoch 8/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.9467 - mae: 7.9467\n",
            "Epoch 9/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 7.1472 - mae: 7.1472\n",
            "Epoch 10/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.7799 - mae: 6.7799\n",
            "Epoch 11/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.0930 - mae: 7.0930\n",
            "Epoch 12/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 7.3919 - mae: 7.3919\n",
            "Epoch 13/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 7.7495 - mae: 7.7495\n",
            "Epoch 14/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 7.8384 - mae: 7.8384\n",
            "Epoch 15/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 7.7167 - mae: 7.7167\n",
            "Epoch 16/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 7.4379 - mae: 7.4379\n",
            "Epoch 17/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 7.0412 - mae: 7.0412\n",
            "Epoch 18/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 6.7393 - mae: 6.7393\n",
            "Epoch 19/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 6.4253 - mae: 6.4253\n",
            "Epoch 20/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.1335 - mae: 6.1335\n",
            "Epoch 21/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 6.0891 - mae: 6.0891\n",
            "Epoch 22/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 6.1179 - mae: 6.1179\n",
            "Epoch 23/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.2186 - mae: 6.2186\n",
            "Epoch 24/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.2103 - mae: 6.2103\n",
            "Epoch 25/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 6.1069 - mae: 6.1069\n",
            "Epoch 26/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.9193 - mae: 5.9193\n",
            "Epoch 27/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.6555 - mae: 5.6555\n",
            "Epoch 28/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.5446 - mae: 5.5446\n",
            "Epoch 29/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 5.4382 - mae: 5.4382\n",
            "Epoch 30/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4606 - mae: 5.4606\n",
            "Epoch 31/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.4736 - mae: 5.4736\n",
            "Epoch 32/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 5.4398 - mae: 5.4398\n",
            "Epoch 33/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.3630 - mae: 5.3630\n",
            "Epoch 34/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.2471 - mae: 5.2471\n",
            "Epoch 35/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 5.0956 - mae: 5.0956\n",
            "Epoch 36/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 4.9120 - mae: 4.9120\n",
            "Epoch 37/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 4.6995 - mae: 4.6995\n",
            "Epoch 38/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 4.6149 - mae: 4.6149\n",
            "Epoch 39/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 4.5299 - mae: 4.5299\n",
            "Epoch 40/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 4.5551 - mae: 4.5551\n",
            "Epoch 41/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 4.4958 - mae: 4.4958\n",
            "Epoch 42/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 4.3239 - mae: 4.3239\n",
            "Epoch 43/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 4.0615 - mae: 4.0615\n",
            "Epoch 44/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.9147 - mae: 3.9147\n",
            "Epoch 45/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.7620 - mae: 3.7620\n",
            "Epoch 46/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.6785 - mae: 3.6785\n",
            "Epoch 47/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.5858 - mae: 3.5858\n",
            "Epoch 48/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.4423 - mae: 3.4423\n",
            "Epoch 49/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.2510 - mae: 3.2510\n",
            "Epoch 50/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 3.0145 - mae: 3.0145\n",
            "Epoch 51/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.8590 - mae: 2.8590\n",
            "Epoch 52/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 2.7603 - mae: 2.7603\n",
            "Epoch 53/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 2.6090 - mae: 2.6090\n",
            "Epoch 54/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 2.3241 - mae: 2.3241\n",
            "Epoch 55/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 2.1140 - mae: 2.1140\n",
            "Epoch 56/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 1.9254 - mae: 1.9254\n",
            "Epoch 57/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 1.7377 - mae: 1.7377\n",
            "Epoch 58/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 1.4872 - mae: 1.4872\n",
            "Epoch 59/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 1.2463 - mae: 1.2463\n",
            "Epoch 60/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 1.1122 - mae: 1.1122\n",
            "Epoch 61/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.8237 - mae: 0.8237\n",
            "Epoch 62/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 0.5397 - mae: 0.5397\n",
            "Epoch 63/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.4359 - mae: 0.4359\n",
            "Epoch 64/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.0797 - mae: 0.0797\n",
            "Epoch 65/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5303 - mae: 0.5303\n",
            "Epoch 66/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.6773 - mae: 0.6773\n",
            "Epoch 67/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5536 - mae: 0.5536\n",
            "Epoch 68/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.7551 - mae: 0.7551\n",
            "Epoch 69/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.8677 - mae: 0.8677\n",
            "Epoch 70/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.7751 - mae: 0.7751\n",
            "Epoch 71/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.5703 - mae: 0.5703\n",
            "Epoch 72/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.5886 - mae: 0.5886\n",
            "Epoch 73/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.4459 - mae: 0.4459\n",
            "Epoch 74/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.3851 - mae: 0.3851\n",
            "Epoch 75/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.2976 - mae: 0.2976\n",
            "Epoch 76/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.0601 - mae: 0.0601\n",
            "Epoch 77/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.4160 - mae: 0.4160\n",
            "Epoch 78/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.5814 - mae: 0.5814\n",
            "Epoch 79/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.5456 - mae: 0.5456\n",
            "Epoch 80/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.3734 - mae: 0.3734\n",
            "Epoch 81/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.6032 - mae: 0.6032\n",
            "Epoch 82/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.7461 - mae: 0.7461\n",
            "Epoch 83/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.6199 - mae: 0.6199\n",
            "Epoch 84/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.2987 - mae: 0.2987\n",
            "Epoch 85/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.4323 - mae: 0.4323\n",
            "Epoch 86/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5550 - mae: 0.5550\n",
            "Epoch 87/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.4488 - mae: 0.4488\n",
            "Epoch 88/100\n",
            "1/1 [==============================] - 0s 9ms/step - loss: 0.0713 - mae: 0.0713\n",
            "Epoch 89/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5674 - mae: 0.5674\n",
            "Epoch 90/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.8465 - mae: 0.8465\n",
            "Epoch 91/100\n",
            "1/1 [==============================] - 0s 11ms/step - loss: 0.8103 - mae: 0.8103\n",
            "Epoch 92/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5187 - mae: 0.5187\n",
            "Epoch 93/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.2017 - mae: 0.2017\n",
            "Epoch 94/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.4824 - mae: 0.4824\n",
            "Epoch 95/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.5530 - mae: 0.5530\n",
            "Epoch 96/100\n",
            "1/1 [==============================] - 0s 13ms/step - loss: 0.3501 - mae: 0.3501\n",
            "Epoch 97/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.1435 - mae: 0.1435\n",
            "Epoch 98/100\n",
            "1/1 [==============================] - 0s 15ms/step - loss: 0.3378 - mae: 0.3378\n",
            "Epoch 99/100\n",
            "1/1 [==============================] - 0s 10ms/step - loss: 0.2840 - mae: 0.2840\n",
            "Epoch 100/100\n",
            "1/1 [==============================] - 0s 12ms/step - loss: 0.1326 - mae: 0.1326\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba289869090>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ],
      "source": [
        "# Let's see if we can make another to improve our model\n",
        "\n",
        "# 1. Create the model (this time with an extra hidden layer with 100 hidden units)\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(50, activation=None),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model.compile(loss=\"mae\",\n",
        "              optimizer=tf.keras.optimizers.Adam(learning_rate=0.01),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model.fit(tf.expand_dims(X, axis=-1), y, epochs=100)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vgAMXNxmOk99",
        "outputId": "59eabbff-679c-4579-abdd-be750543f812"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(8,), dtype=float32, numpy=array([-7., -4., -1.,  2.,  5.,  8., 11., 14.], dtype=float32)>,\n",
              " <tf.Tensor: shape=(8,), dtype=float32, numpy=array([ 3.,  6.,  9., 12., 15., 18., 21., 24.], dtype=float32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ],
      "source": [
        "# Let's remind ourselves of the data\n",
        "X, y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BvsC0lWoOPl0",
        "outputId": "76c869c2-e0d1-40e3-983c-b3f1291c7aa5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 69ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[26.573]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 20
        }
      ],
      "source": [
        "# Let's try to make a prediction\n",
        "model.predict([17.0])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ReQcDQVCObIr"
      },
      "source": [
        "## Evaluating a TensorFlow model\n",
        "\n",
        "In practice, a typical workflow you'll go through when building neural networks is:\n",
        "\n",
        "```\n",
        "Build a model -> fit it -> evaluate it -> tweak a model -> fit it -> evaluate it -> tweak a model -> fit it -> evaluate it...\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NsIu7y-ze5EM"
      },
      "source": [
        "### Part 1: \"Visualize, visualize, visualize\"\n",
        "\n",
        "When it comes to evaluation... there are 3 words you should memorize:\n",
        "> \"Visualize, visualize, visualize\"\n",
        "\n",
        "It's a good idea to visualize:\n",
        "*\n",
        "* The data - What data awre we working with? What does it look like?\n",
        "* The model itself - What does our model look like?\n",
        "* The training of a model - How does it perform while it learns?\n",
        "* The predictions of the model - how do the predictions of a model line up against the ground truth (the original labels)?"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jPr8C_I9fySj",
        "outputId": "ea7bbf8a-88fb-4cca-d352-3c33a5a7bdf4"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(50,), dtype=int32, numpy=\n",
              "array([-100,  -96,  -92,  -88,  -84,  -80,  -76,  -72,  -68,  -64,  -60,\n",
              "        -56,  -52,  -48,  -44,  -40,  -36,  -32,  -28,  -24,  -20,  -16,\n",
              "        -12,   -8,   -4,    0,    4,    8,   12,   16,   20,   24,   28,\n",
              "         32,   36,   40,   44,   48,   52,   56,   60,   64,   68,   72,\n",
              "         76,   80,   84,   88,   92,   96], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ],
      "source": [
        "# Make a bigger dataset\n",
        "X = tf.range(-100, 100, 4)\n",
        "X"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ituq5PgVgUsk",
        "outputId": "22aca66b-4ff2-40a1-830f-5baec2219ab5"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(50,), dtype=int32, numpy=\n",
              "array([-90, -86, -82, -78, -74, -70, -66, -62, -58, -54, -50, -46, -42,\n",
              "       -38, -34, -30, -26, -22, -18, -14, -10,  -6,  -2,   2,   6,  10,\n",
              "        14,  18,  22,  26,  30,  34,  38,  42,  46,  50,  54,  58,  62,\n",
              "        66,  70,  74,  78,  82,  86,  90,  94,  98, 102, 106], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ],
      "source": [
        "# Make labels for the dataset\n",
        "y = X + 10\n",
        "y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 430
        },
        "id": "1J5K75kQgXme",
        "outputId": "0abaf417-e194-4b88-8627-e84fe73bd8f8"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Visualize the data\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.scatter(X, y);"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7nY7JNviggxU"
      },
      "source": [
        "### Part 2: The 3 datasets...\n",
        "\n",
        "* **Training set** - the model learns from this data, which is typically 70-80% of the total data you have available.\n",
        "* **Validation set** - The model gets tuned on this data, which is typically 10-15% of the data available.\n",
        "* **Test set** - The model gets evaluated on this data to test what it has learned. This set is typically 10-15% of the total data available."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EALXVtbczH4o",
        "outputId": "d7c1759b-194a-4ceb-8abb-e6eac6670805"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "50"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ],
      "source": [
        "# Check the length of how many samples we have\n",
        "len(X)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7CVZVUvhzN-K",
        "outputId": "c55552a9-ed1f-42be-c6c8-856cd008ce62"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(40, 10, 40, 10)"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ],
      "source": [
        "# Split the data into train and test sets\n",
        "X_train = X[:40] # first 40 are training samples (80% of the data)\n",
        "y_train = y[:40]\n",
        "\n",
        "X_test = X[40:] # last 10 are testing samples (20% of the data)\n",
        "y_test = y[40:]\n",
        "\n",
        "len(X_train), len(X_test), len(y_train), len(y_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cSAp1549z0XO"
      },
      "source": [
        "#### Visualizing the data\n",
        "Now we've got our data in training and test sets... let's visualize it again!"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 599
        },
        "id": "VCuyEfPR0OCw",
        "outputId": "583d65d4-fca9-4582-bfae-4d7cbde5385b"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plt.figure(figsize=(10, 7))\n",
        "# Plot training data in blue\n",
        "plt.scatter(X_train, y_train, c=\"b\", label=\"Training data\")\n",
        "# Plot test data in green\n",
        "plt.scatter(X_test, y_test, c=\"g\", label=\"Testing data\")\n",
        "# Show a legend\n",
        "plt.legend();"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_kBuVQ9e0b16"
      },
      "source": [
        "### Part 3: Getting a model summary"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "5ZcTe3ZK1H5o"
      },
      "outputs": [],
      "source": [
        "# Let's have a look on how to build a neural network for our data\n",
        "\n",
        "# 1. Create a model\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "# model.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=100)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MF61YbnP1rHx"
      },
      "source": [
        "#### Visualizing the model\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eDZOu7e-2Ca2"
      },
      "outputs": [],
      "source": [
        "# model.summary() # Wont' work if not built already with build() or fit()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SZKouAW02yOL",
        "outputId": "835d704c-0282-45f7-c927-7da41ea1438d"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(), dtype=int32, numpy=-100>,\n",
              " <tf.Tensor: shape=(), dtype=int32, numpy=-90>)"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ],
      "source": [
        "X[0], y[0] # they are scalars"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8JCXuJ9x2NSQ"
      },
      "outputs": [],
      "source": [
        "# Let's create a model which builds automatically by defining the input_shape argument in the first layer.\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create a model (same as above)\n",
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1, input_shape=[1]), # define the input shape of the model\n",
        "], name=\"model_1\")\n",
        "\n",
        "# 2. Compile the model\n",
        "model.compile(loss=tf.keras.losses.mae,\n",
        "              optimizer=tf.keras.optimizers.SGD(),\n",
        "              metrics=[\"mae\"])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "14cbjGeL3TbB"
      },
      "source": [
        "* Total params - total number of parameters in the model.\n",
        "* Trainable params - these are the parameters (patterns) the model can update as it trains.\n",
        "* Non-trainable params - these parameters aren't updated during training (this is typical when you bring in already learnt patterns or parameters from other models during **transfer learning**).\n",
        "\n",
        "**Resource:** For a more in-depth overview of the trainable parameters within a layer, check out MIT's introduction to deep learning video. http://introtodeeplearning.com/\n",
        "\n",
        "**Exercise:** Try playing around with the number of hidden units in the dense layer, see how that effects the number of parameters (total and trainable) by calling `model.summary()`."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "exbQjD3E4HEw",
        "outputId": "49e01771-1c40-4877-9ba2-8a37c83c8b73"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba2880c7520>"
            ]
          },
          "metadata": {},
          "execution_count": 31
        }
      ],
      "source": [
        "# Let's fit our model to the training data\n",
        "model.fit(X_train, y_train, epochs=100, verbose=0)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "woE4RFgn6HZo"
      },
      "source": [
        "### Part 4: Visualizing a model's layers"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iyWp_YZM-_rY",
        "outputId": "951b6f99-b8f2-4056-a9ba-620fbfff2b27"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_1\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_5 (Dense)             (None, 1)                 2         \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 2 (8.00 Byte)\n",
            "Trainable params: 2 (8.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "# Get a summary of our model\n",
        "model.summary()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 201
        },
        "id": "bGrirpBi_CvI",
        "outputId": "a0cd2616-f097-4679-c658-a46d5b56a11e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ],
      "source": [
        "from tensorflow.keras.utils import plot_model\n",
        "\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/utils/plot_model\n",
        "plot_model(model=model, show_shapes=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P93yTUFH_opI"
      },
      "source": [
        "### Part 5: Visualizing our model's predictions\n",
        "\n",
        "To visualize predictions, it's a good idea to plot them against the ground truth labels.\n",
        "\n",
        "Often you'll see this in the form of `y_test` or `y_true` versus `y_pred` (ground truth versus your model's predictions)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pCisY8jqBXwN",
        "outputId": "f6785269-cd40-4f3e-dea2-ff51b7bfb70c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 46ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[44.504948],\n",
              "       [47.385445],\n",
              "       [50.26594 ],\n",
              "       [53.14644 ],\n",
              "       [56.026936],\n",
              "       [58.907433],\n",
              "       [61.787926],\n",
              "       [64.66842 ],\n",
              "       [67.54892 ],\n",
              "       [70.42941 ]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ],
      "source": [
        "# Make some predictions\n",
        "y_pred = model.predict(X_test)\n",
        "y_pred"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Cv1ZkLB3CARd",
        "outputId": "e03a63da-1288-4f0b-9022-50203a7e7f18"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=int32, numpy=array([ 70,  74,  78,  82,  86,  90,  94,  98, 102, 106], dtype=int32)>"
            ]
          },
          "metadata": {},
          "execution_count": 35
        }
      ],
      "source": [
        "y_test"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OmSyU-1QCLWX"
      },
      "source": [
        "**Note:** If you feel like you're going to reuse some kind of functionality in the future, it's a good idea to turn it into a function."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gpYAWF4tCDT7"
      },
      "outputs": [],
      "source": [
        "# Let's create a plotting function\n",
        "def plot_predictions(train_data=X_train,\n",
        "                     train_labels=y_train,\n",
        "                     test_data=X_test,\n",
        "                     test_labels=y_test,\n",
        "                     predictions=y_pred):\n",
        "  \"\"\"\n",
        "  Plots training data, test data and compares predictions to ground truth labels.\n",
        "  \"\"\"\n",
        "  plt.figure(figsize=(10, 7))\n",
        "  # Plot training data in blue\n",
        "  plt.scatter(train_data, train_labels, c=\"b\", label=\"Training data\")\n",
        "  # Plot testing data in green\n",
        "  plt.scatter(test_data, test_labels, c=\"g\", label=\"Testing data\")\n",
        "  # Plot model's predictions in red\n",
        "  plt.scatter(test_data, predictions, c=\"r\", label=\"Predictions\")\n",
        "  # Show the legend\n",
        "  plt.legend();"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 599
        },
        "id": "_frkkHowDHb_",
        "outputId": "19813d0f-b2e8-4e9e-a193-1ca5774fe0be"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "plot_predictions(train_data=X_train,\n",
        "                 test_data=X_test,\n",
        "                 train_labels=y_train,\n",
        "                 test_labels=y_test,\n",
        "                 predictions=y_pred)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ygcTa06WDJou"
      },
      "source": [
        "### Part 6: Common regression evaluation metrics\n",
        "\n",
        "Depending on the problem you're working on, there will be differnet evaluation metrics to evaluate your model's performance.\n",
        "\n",
        "Since we're working on a regression, two of the main metrics:\n",
        "* MAE - mean absolute error, \"on average, how wrong is each of my model's predictions\"\n",
        "* MSE - mean square error, \"square the average errors\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xcGlpWnpEoaQ",
        "outputId": "b9d1c84a-d12f-4a4c-f1a3-01ebbc699278"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 151ms/step - loss: 30.5328 - mae: 30.5328\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[30.532817840576172, 30.532817840576172]"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ],
      "source": [
        "# Evaluate the model on the test set\n",
        "model.evaluate(X_test, y_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UzwRtiwYGfQn"
      },
      "source": [
        "### Part 7a: Mean absolute error"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SR41190ZGNb5",
        "outputId": "39fc08f3-a664-4466-df5c-9b40f48b223e"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=float32, numpy=\n",
              "array([43.495052, 40.614555, 37.73406 , 34.85356 , 31.973064, 29.092566,\n",
              "       26.212072, 23.331581, 20.45108 , 17.65647 ], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ],
      "source": [
        "# Calculate the mean absolute error\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/losses/MeanAbsoluteError\n",
        "mae = tf.metrics.mean_absolute_error(y_true=y_test,\n",
        "                                     y_pred=y_pred)\n",
        "mae"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "99Uc7mweGL_A",
        "outputId": "400ef770-7612-4dcb-91ce-82623bb48c0a"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(10, 1), dtype=float32, numpy=\n",
              " array([[44.504948],\n",
              "        [47.385445],\n",
              "        [50.26594 ],\n",
              "        [53.14644 ],\n",
              "        [56.026936],\n",
              "        [58.907433],\n",
              "        [61.787926],\n",
              "        [64.66842 ],\n",
              "        [67.54892 ],\n",
              "        [70.42941 ]], dtype=float32)>,\n",
              " <tf.Tensor: shape=(10,), dtype=int32, numpy=array([ 70,  74,  78,  82,  86,  90,  94,  98, 102, 106], dtype=int32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ],
      "source": [
        "tf.constant(y_pred), y_test"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zu7uq5CdHCzo",
        "outputId": "8a807063-3e6d-4593-cc50-d605e75c6dff"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(10,), dtype=float32, numpy=\n",
              "array([44.504948, 47.385445, 50.26594 , 53.14644 , 56.026936, 58.907433,\n",
              "       61.787926, 64.66842 , 67.54892 , 70.42941 ], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 41
        }
      ],
      "source": [
        "tf.squeeze(y_pred)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rfp2_XA1HWAy",
        "outputId": "0943b0bd-86be-4030-8c14-d17195415736"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=30.532818>"
            ]
          },
          "metadata": {},
          "execution_count": 42
        }
      ],
      "source": [
        "# Calculate the mean absolute error using reshaped y_pred\n",
        "mae = tf.metrics.mean_absolute_error(y_true=y_test,\n",
        "                                     y_pred=tf.squeeze(y_pred))\n",
        "mae"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hV8niLn3HcKU"
      },
      "source": [
        "### Part 7b: Mean squared error"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mejUddqyJULM",
        "outputId": "3f9bc5c8-6be0-45d1-b53d-24cc2a9f5f2c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(), dtype=float32, numpy=942.5926>"
            ]
          },
          "metadata": {},
          "execution_count": 43
        }
      ],
      "source": [
        "# Calculate the mean squared error\n",
        "# https://www.tensorflow.org/api_docs/python/tf/keras/losses/MeanSquaredError\n",
        "mse = tf.metrics.mean_squared_error(y_true=y_test,\n",
        "                                    y_pred=tf.squeeze(y_pred))\n",
        "mse"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vXN8Ys5LJjl_"
      },
      "outputs": [],
      "source": [
        "# Make some functions to reuse MAE and MSE\n",
        "def mae(y_true, y_pred):\n",
        "  return tf.metrics.mean_absolute_error(y_true=y_true,\n",
        "                                        y_pred=tf.squeeze(y_pred))\n",
        "\n",
        "def mse(y_true, y_pred):\n",
        "  return tf.metrics.mean_squared_error(y_true=y_true,\n",
        "                                       y_pred=tf.squeeze(y_pred))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NYTdZIaBKDsG"
      },
      "source": [
        "## Setting up TensorFlow modelling experiments to improve our model\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "79EWfKDNUw84"
      },
      "source": [
        "### Part 1: Start with a simple model\n",
        "```\n",
        "Build a model -> fit it -> evaluate it -> tweak a model -> fit it -> evaluate it -> tweak a model -> fit it -> evaluate it...\n",
        "```\n",
        "\n",
        "1. Get more data - get more examples for your model to train on (more opportunities to learn patterns or relationships between features and labels).\n",
        "2. Make your model larger (using a more complex model) - this might come in the form of more layers or more hidden units in each layer.\n",
        "3. Train for longer - give your modelmore of a chance to find patterns in the data.\n",
        "\n",
        "Let's do 3 modelling experiments:\n",
        "1. `model_1` - same as the original model, 1 layer, trained for 100 epochs.\n",
        "2. `model_2` - 2 layers, trained for 100 epochs\n",
        "3. `model_3` - 2 layers, trained for 500 epochs"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1m79r14oUzzM"
      },
      "source": [
        "**Build** `model_1`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rIsEiy8JWDdH",
        "outputId": "b6a2bd60-f927-40fe-866f-815a4883192c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "2/2 [==============================] - 0s 23ms/step - loss: 84.0947 - mae: 84.0947\n",
            "Epoch 2/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 48.6392 - mae: 48.6392\n",
            "Epoch 3/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.2219 - mae: 14.2219\n",
            "Epoch 4/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 7.6748 - mae: 7.6748\n",
            "Epoch 5/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.1922 - mae: 9.1922\n",
            "Epoch 6/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.6568 - mae: 8.6568\n",
            "Epoch 7/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.0821 - mae: 9.0821\n",
            "Epoch 8/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.1112 - mae: 9.1112\n",
            "Epoch 9/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 19.9861 - mae: 19.9861\n",
            "Epoch 10/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.8186 - mae: 10.8186\n",
            "Epoch 11/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.7044 - mae: 8.7044\n",
            "Epoch 12/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.6855 - mae: 9.6855\n",
            "Epoch 13/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.0686 - mae: 13.0686\n",
            "Epoch 14/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.1590 - mae: 14.1590\n",
            "Epoch 15/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.5619 - mae: 11.5619\n",
            "Epoch 16/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.5242 - mae: 8.5242\n",
            "Epoch 17/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.5708 - mae: 13.5708\n",
            "Epoch 18/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3761 - mae: 11.3761\n",
            "Epoch 19/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.0718 - mae: 18.0718\n",
            "Epoch 20/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.2053 - mae: 15.2053\n",
            "Epoch 21/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.1582 - mae: 11.1582\n",
            "Epoch 22/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2849 - mae: 8.2849\n",
            "Epoch 23/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.4648 - mae: 9.4648\n",
            "Epoch 24/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 7.7145 - mae: 7.7145\n",
            "Epoch 25/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.0223 - mae: 13.0223\n",
            "Epoch 26/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.5747 - mae: 16.5747\n",
            "Epoch 27/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.2985 - mae: 13.2985\n",
            "Epoch 28/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.4204 - mae: 14.4204\n",
            "Epoch 29/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.1330 - mae: 10.1330\n",
            "Epoch 30/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.5697 - mae: 16.5697\n",
            "Epoch 31/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.4572 - mae: 23.4572\n",
            "Epoch 32/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.3926 - mae: 7.3926\n",
            "Epoch 33/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.8215 - mae: 9.8215\n",
            "Epoch 34/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 12.2642 - mae: 12.2642\n",
            "Epoch 35/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.1958 - mae: 11.1958\n",
            "Epoch 36/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.4273 - mae: 13.4273\n",
            "Epoch 37/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.4674 - mae: 9.4674\n",
            "Epoch 38/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.1187 - mae: 10.1187\n",
            "Epoch 39/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.9950 - mae: 8.9950\n",
            "Epoch 40/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.6237 - mae: 9.6237\n",
            "Epoch 41/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.5831 - mae: 10.5831\n",
            "Epoch 42/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.6363 - mae: 10.6363\n",
            "Epoch 43/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.2357 - mae: 7.2357\n",
            "Epoch 44/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.0361 - mae: 8.0361\n",
            "Epoch 45/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.8380 - mae: 9.8380\n",
            "Epoch 46/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.9165 - mae: 8.9165\n",
            "Epoch 47/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.5921 - mae: 7.5921\n",
            "Epoch 48/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.5621 - mae: 8.5621\n",
            "Epoch 49/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.0487 - mae: 10.0487\n",
            "Epoch 50/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.0089 - mae: 9.0089\n",
            "Epoch 51/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7243 - mae: 10.7243\n",
            "Epoch 52/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.2861 - mae: 15.2861\n",
            "Epoch 53/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.3128 - mae: 14.3128\n",
            "Epoch 54/100\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 21.5669 - mae: 21.5669\n",
            "Epoch 55/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.0511 - mae: 16.0511\n",
            "Epoch 56/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.2700 - mae: 10.2700\n",
            "Epoch 57/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.8139 - mae: 9.8139\n",
            "Epoch 58/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.0975 - mae: 9.0975\n",
            "Epoch 59/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.2987 - mae: 8.2987\n",
            "Epoch 60/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.3945 - mae: 9.3945\n",
            "Epoch 61/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.2204 - mae: 11.2204\n",
            "Epoch 62/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.0575 - mae: 12.0575\n",
            "Epoch 63/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.2942 - mae: 7.2942\n",
            "Epoch 64/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 12.4762 - mae: 12.4762\n",
            "Epoch 65/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.5563 - mae: 10.5563\n",
            "Epoch 66/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.5884 - mae: 15.5884\n",
            "Epoch 67/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.0043 - mae: 10.0043\n",
            "Epoch 68/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.7327 - mae: 8.7327\n",
            "Epoch 69/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.4663 - mae: 13.4663\n",
            "Epoch 70/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.4933 - mae: 7.4933\n",
            "Epoch 71/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.2832 - mae: 12.2832\n",
            "Epoch 72/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.5332 - mae: 8.5332\n",
            "Epoch 73/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.0599 - mae: 7.0599\n",
            "Epoch 74/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.9117 - mae: 9.9117\n",
            "Epoch 75/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.9826 - mae: 9.9826\n",
            "Epoch 76/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.1433 - mae: 10.1433\n",
            "Epoch 77/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.0102 - mae: 13.0102\n",
            "Epoch 78/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.9146 - mae: 10.9146\n",
            "Epoch 79/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.3893 - mae: 15.3893\n",
            "Epoch 80/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.7548 - mae: 11.7548\n",
            "Epoch 81/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2661 - mae: 9.2661\n",
            "Epoch 82/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.7507 - mae: 12.7507\n",
            "Epoch 83/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.3064 - mae: 8.3064\n",
            "Epoch 84/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.3276 - mae: 7.3276\n",
            "Epoch 85/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4846 - mae: 10.4846\n",
            "Epoch 86/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.1590 - mae: 9.1590\n",
            "Epoch 87/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.8666 - mae: 11.8666\n",
            "Epoch 88/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4769 - mae: 10.4769\n",
            "Epoch 89/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.9925 - mae: 6.9925\n",
            "Epoch 90/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.8001 - mae: 13.8001\n",
            "Epoch 91/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 7.8857 - mae: 7.8857\n",
            "Epoch 92/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.9365 - mae: 7.9365\n",
            "Epoch 93/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.4319 - mae: 9.4319\n",
            "Epoch 94/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.7529 - mae: 8.7529\n",
            "Epoch 95/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.8695 - mae: 9.8695\n",
            "Epoch 96/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 7.3633 - mae: 7.3633\n",
            "Epoch 97/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.8774 - mae: 6.8774\n",
            "Epoch 98/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.2710 - mae: 7.2710\n",
            "Epoch 99/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.4228 - mae: 7.4228\n",
            "Epoch 100/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.1255 - mae: 11.1255\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba28835c730>"
            ]
          },
          "metadata": {},
          "execution_count": 45
        }
      ],
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create the model\n",
        "model_1 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model_1.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model_1.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=100)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make and plot predictions for model_1\n",
        "y_preds_1 = model_1.predict(X_test)\n",
        "plot_predictions(predictions=y_preds_1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 654
        },
        "id": "amvhnii4c3MI",
        "outputId": "83e280b4-f671-4142-8192-d8f09ad3e47e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:tensorflow:5 out of the last 5 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7ba288132320> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 46ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tf.constant(y_preds_1), tf.squeeze(y_preds_1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uHBIt3iPeWr4",
        "outputId": "63261186-ac7e-4777-bd26-37e993d60cc3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(10, 1), dtype=float32, numpy=\n",
              " array([[51.574207],\n",
              "        [54.926945],\n",
              "        [58.279682],\n",
              "        [61.632423],\n",
              "        [64.98516 ],\n",
              "        [68.337906],\n",
              "        [71.69064 ],\n",
              "        [75.04338 ],\n",
              "        [78.39612 ],\n",
              "        [81.748856]], dtype=float32)>,\n",
              " <tf.Tensor: shape=(10,), dtype=float32, numpy=\n",
              " array([51.574207, 54.926945, 58.279682, 61.632423, 64.98516 , 68.337906,\n",
              "        71.69064 , 75.04338 , 78.39612 , 81.748856], dtype=float32)>)"
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate model_1 evaluation metrics\n",
        "mae_1 = mae(y_test, y_preds_1)\n",
        "mse_1 = mse(y_test, y_preds_1)\n",
        "mae_1, mse_1"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6UF62vh9d308",
        "outputId": "e6cc33f3-6599-4564-9f90-beda44a7f818"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(), dtype=float32, numpy=21.338467>,\n",
              " <tf.Tensor: shape=(), dtype=float32, numpy=458.78653>)"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Part 2: Increasing complexity"
      ],
      "metadata": {
        "id": "2geBxE8bfOsM"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Build `model_2`**\n",
        "\n",
        "* 2 Dense (fully connected) layers, trained for 100 epochs"
      ],
      "metadata": {
        "id": "TTxBM6dveN9p"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Set the random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create the model\n",
        "model_2 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model_2.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mse\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model_2.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=100)"
      ],
      "metadata": {
        "id": "L5ntDEGXfGYC",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "663dcefd-a726-4561-e892-1929af74e258"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 45.3275 - mse: 2850.1848\n",
            "Epoch 2/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 33.5377 - mse: 1482.6449\n",
            "Epoch 3/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 37.1217 - mse: 2136.1372\n",
            "Epoch 4/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 23.2199 - mse: 773.1878\n",
            "Epoch 5/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.9323 - mse: 241.0312\n",
            "Epoch 6/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.2488 - mse: 141.2601\n",
            "Epoch 7/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.3248 - mse: 196.0737\n",
            "Epoch 8/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 11.0181 - mse: 141.0269\n",
            "Epoch 9/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 39.1048 - mse: 2379.6899\n",
            "Epoch 10/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 26.6777 - mse: 1000.3781\n",
            "Epoch 11/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.8091 - mse: 219.4995\n",
            "Epoch 12/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.6125 - mse: 790.7162\n",
            "Epoch 13/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 21.2857 - mse: 646.8953\n",
            "Epoch 14/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.7743 - mse: 867.5082\n",
            "Epoch 15/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 15.8185 - mse: 340.9097\n",
            "Epoch 16/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4538 - mse: 124.9623\n",
            "Epoch 17/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 19.7470 - mse: 531.4861\n",
            "Epoch 18/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.3629 - mse: 280.3950\n",
            "Epoch 19/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.3788 - mse: 430.4216\n",
            "Epoch 20/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2480 - mse: 93.4026\n",
            "Epoch 21/100\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 14.4564 - mse: 289.5242\n",
            "Epoch 22/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.8744 - mse: 246.6252\n",
            "Epoch 23/100\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 15.5062 - mse: 315.5447\n",
            "Epoch 24/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 15.2964 - mse: 316.8399\n",
            "Epoch 25/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.3583 - mse: 273.0024\n",
            "Epoch 26/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.3796 - mse: 569.8639\n",
            "Epoch 27/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.4324 - mse: 167.5402\n",
            "Epoch 28/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 28.9870 - mse: 1365.2307\n",
            "Epoch 29/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2485 - mse: 94.7453\n",
            "Epoch 30/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.7480 - mse: 1600.4288\n",
            "Epoch 31/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 53.9665 - mse: 5216.8213\n",
            "Epoch 32/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.5480 - mse: 100.9348\n",
            "Epoch 33/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.1368 - mse: 179.5043\n",
            "Epoch 34/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.8023 - mse: 858.0106\n",
            "Epoch 35/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.6360 - mse: 242.0522\n",
            "Epoch 36/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 21.5658 - mse: 663.3178\n",
            "Epoch 37/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3468 - mse: 149.3588\n",
            "Epoch 38/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.4118 - mse: 266.5790\n",
            "Epoch 39/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7640 - mse: 137.9282\n",
            "Epoch 40/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 16.5269 - mse: 396.5495\n",
            "Epoch 41/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.9373 - mse: 179.9495\n",
            "Epoch 42/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2743 - mse: 115.3718\n",
            "Epoch 43/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.5840 - mse: 111.2997\n",
            "Epoch 44/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 27.9006 - mse: 1247.7917\n",
            "Epoch 45/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.2585 - mse: 146.1529\n",
            "Epoch 46/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 13.8524 - mse: 280.8486\n",
            "Epoch 47/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.9214 - mse: 211.9884\n",
            "Epoch 48/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.9553 - mse: 389.3326\n",
            "Epoch 49/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.7724 - mse: 104.8677\n",
            "Epoch 50/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 14.1656 - mse: 280.7806\n",
            "Epoch 51/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.7348 - mse: 161.2185\n",
            "Epoch 52/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 31.3942 - mse: 1712.5808\n",
            "Epoch 53/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.7268 - mse: 321.2610\n",
            "Epoch 54/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 24.6203 - mse: 912.1613\n",
            "Epoch 55/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.9039 - mse: 859.1110\n",
            "Epoch 56/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.1218 - mse: 171.0175\n",
            "Epoch 57/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.0633 - mse: 211.6613\n",
            "Epoch 58/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.7826 - mse: 105.6529\n",
            "Epoch 59/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.2465 - mse: 245.9900\n",
            "Epoch 60/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.8377 - mse: 138.0863\n",
            "Epoch 61/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 13.4244 - mse: 238.8992\n",
            "Epoch 62/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.4083 - mse: 462.1849\n",
            "Epoch 63/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.1407 - mse: 87.6368\n",
            "Epoch 64/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 18.2841 - mse: 486.0117\n",
            "Epoch 65/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.0962 - mse: 114.1039\n",
            "Epoch 66/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 24.1276 - mse: 877.4763\n",
            "Epoch 67/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.8516 - mse: 141.1431\n",
            "Epoch 68/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7443 - mse: 156.7581\n",
            "Epoch 69/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 23.1092 - mse: 771.3634\n",
            "Epoch 70/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.8394 - mse: 142.9906\n",
            "Epoch 71/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.0644 - mse: 351.9764\n",
            "Epoch 72/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.7551 - mse: 91.3568\n",
            "Epoch 73/100\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 10.1589 - mse: 165.3297\n",
            "Epoch 74/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 28.3361 - mse: 1123.7031\n",
            "Epoch 75/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.0062 - mse: 146.1944\n",
            "Epoch 76/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.9219 - mse: 207.6815\n",
            "Epoch 77/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.8826 - mse: 498.3127\n",
            "Epoch 78/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.9630 - mse: 87.7962\n",
            "Epoch 79/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 28.7098 - mse: 1301.6163\n",
            "Epoch 80/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 31.0465 - mse: 1530.4851\n",
            "Epoch 81/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.4130 - mse: 251.6239\n",
            "Epoch 82/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.3236 - mse: 284.4959\n",
            "Epoch 83/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.8604 - mse: 471.9992\n",
            "Epoch 84/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.4610 - mse: 149.1018\n",
            "Epoch 85/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 25.2533 - mse: 892.7289\n",
            "Epoch 86/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 16.4214 - mse: 396.6288\n",
            "Epoch 87/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.4682 - mse: 156.3062\n",
            "Epoch 88/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 27.0267 - mse: 1044.7708\n",
            "Epoch 89/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.7414 - mse: 128.3324\n",
            "Epoch 90/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.0744 - mse: 521.7433\n",
            "Epoch 91/100\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.5992 - mse: 163.4034\n",
            "Epoch 92/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.2918 - mse: 458.3460\n",
            "Epoch 93/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 6.8641 - mse: 67.9145\n",
            "Epoch 94/100\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.0696 - mse: 232.0467\n",
            "Epoch 95/100\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 18.4424 - mse: 501.8961\n",
            "Epoch 96/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4095 - mse: 179.3403\n",
            "Epoch 97/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.1852 - mse: 302.1564\n",
            "Epoch 98/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.6509 - mse: 137.1321\n",
            "Epoch 99/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3440 - mse: 164.6432\n",
            "Epoch 100/100\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.3061 - mse: 1322.4629\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba288d17ac0>"
            ]
          },
          "metadata": {},
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make and plot predictions of model 2\n",
        "y_preds_2 = model_2.predict(X_test)\n",
        "plot_predictions(predictions=y_preds_2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 654
        },
        "id": "VaE0ACTYhVLm",
        "outputId": "4eed63ae-e922-44fb-89e2-7b16be61da8b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:tensorflow:6 out of the last 6 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7ba288b8c8b0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 71ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate model_2 evaluation metrics\n",
        "mae_2 = mae(y_test, y_preds_2)\n",
        "mse_2 = mse(y_test, y_preds_2)\n",
        "mae_2, mse_2"
      ],
      "metadata": {
        "id": "bLAK4B0jhlWs",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c2b3e8e2-b3a8-4f39-b9c3-731693990397"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(), dtype=float32, numpy=11.590208>,\n",
              " <tf.Tensor: shape=(), dtype=float32, numpy=142.92845>)"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Build `model_3`**\n",
        "\n",
        "* 2 layers, trained for 500 epochs"
      ],
      "metadata": {
        "id": "iBOzq6cVh6-h"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create a model\n",
        "model_3 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "model_3.compile(loss=tf.keras.losses.mae,\n",
        "                optimizer=tf.keras.optimizers.SGD(),\n",
        "                metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "model_3.fit(tf.expand_dims(X_train, axis=-1), y_train, epochs=500)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uPSV5lx2iTry",
        "outputId": "911f4018-f630-45d5-e617-bb26f6cb8cbe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 31.6149 - mae: 31.6149\n",
            "Epoch 2/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 26.8223 - mae: 26.8223\n",
            "Epoch 3/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 30.2441 - mae: 30.2441\n",
            "Epoch 4/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.0427 - mae: 16.0427\n",
            "Epoch 5/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 21.7484 - mae: 21.7484\n",
            "Epoch 6/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 11.8379 - mae: 11.8379\n",
            "Epoch 7/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.9687 - mae: 12.9687\n",
            "Epoch 8/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.1988 - mae: 11.1988\n",
            "Epoch 9/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 41.2607 - mae: 41.2607\n",
            "Epoch 10/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 28.5541 - mae: 28.5541\n",
            "Epoch 11/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.4058 - mae: 9.4058\n",
            "Epoch 12/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 26.3904 - mae: 26.3904\n",
            "Epoch 13/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.4664 - mae: 14.4664\n",
            "Epoch 14/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 29.8957 - mae: 29.8957\n",
            "Epoch 15/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 19.7461 - mae: 19.7461\n",
            "Epoch 16/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.8526 - mae: 9.8526\n",
            "Epoch 17/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.6712 - mae: 17.6712\n",
            "Epoch 18/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 14.1424 - mae: 14.1424\n",
            "Epoch 19/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.9832 - mae: 13.9832\n",
            "Epoch 20/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.2244 - mae: 11.2244\n",
            "Epoch 21/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.3034 - mae: 17.3034\n",
            "Epoch 22/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.5411 - mae: 15.5411\n",
            "Epoch 23/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2446 - mae: 9.2446\n",
            "Epoch 24/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.2601 - mae: 17.2601\n",
            "Epoch 25/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.9377 - mae: 15.9377\n",
            "Epoch 26/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 20.9220 - mae: 20.9220\n",
            "Epoch 27/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 25.9842 - mae: 25.9842\n",
            "Epoch 28/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 18.4803 - mae: 18.4803\n",
            "Epoch 29/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.2496 - mae: 9.2496\n",
            "Epoch 30/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 29.0925 - mae: 29.0925\n",
            "Epoch 31/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 52.6937 - mae: 52.6937\n",
            "Epoch 32/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.9358 - mae: 11.9358\n",
            "Epoch 33/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 15.5276 - mae: 15.5276\n",
            "Epoch 34/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.6214 - mae: 12.6214\n",
            "Epoch 35/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2112 - mae: 9.2112\n",
            "Epoch 36/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.5112 - mae: 16.5112\n",
            "Epoch 37/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.0812 - mae: 11.0812\n",
            "Epoch 38/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 18.2095 - mae: 18.2095\n",
            "Epoch 39/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 19.1427 - mae: 19.1427\n",
            "Epoch 40/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 20.5054 - mae: 20.5054\n",
            "Epoch 41/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.8277 - mae: 14.8277\n",
            "Epoch 42/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.2196 - mae: 12.2196\n",
            "Epoch 43/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7058 - mae: 10.7058\n",
            "Epoch 44/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 32.2998 - mae: 32.2998\n",
            "Epoch 45/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 12.4843 - mae: 12.4843\n",
            "Epoch 46/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 17.5420 - mae: 17.5420\n",
            "Epoch 47/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 15.7504 - mae: 15.7504\n",
            "Epoch 48/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.3754 - mae: 8.3754\n",
            "Epoch 49/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.9840 - mae: 13.9840\n",
            "Epoch 50/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 12.8241 - mae: 12.8241\n",
            "Epoch 51/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.8370 - mae: 14.8370\n",
            "Epoch 52/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.7239 - mae: 18.7239\n",
            "Epoch 53/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 24.1676 - mae: 24.1676\n",
            "Epoch 54/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.1826 - mae: 23.1826\n",
            "Epoch 55/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 24.0358 - mae: 24.0358\n",
            "Epoch 56/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.1685 - mae: 11.1685\n",
            "Epoch 57/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.1225 - mae: 13.1225\n",
            "Epoch 58/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.8165 - mae: 9.8165\n",
            "Epoch 59/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.3052 - mae: 13.3052\n",
            "Epoch 60/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.8739 - mae: 10.8739\n",
            "Epoch 61/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.4711 - mae: 13.4711\n",
            "Epoch 62/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.4859 - mae: 17.4859\n",
            "Epoch 63/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.1629 - mae: 9.1629\n",
            "Epoch 64/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.3561 - mae: 18.3561\n",
            "Epoch 65/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.1177 - mae: 10.1177\n",
            "Epoch 66/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 24.2090 - mae: 24.2090\n",
            "Epoch 67/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.8816 - mae: 10.8816\n",
            "Epoch 68/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7670 - mae: 10.7670\n",
            "Epoch 69/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.1844 - mae: 23.1844\n",
            "Epoch 70/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.8304 - mae: 8.8304\n",
            "Epoch 71/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.0061 - mae: 16.0061\n",
            "Epoch 72/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.1616 - mae: 8.1616\n",
            "Epoch 73/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.4720 - mae: 9.4720\n",
            "Epoch 74/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 28.1771 - mae: 28.1771\n",
            "Epoch 75/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.1829 - mae: 10.1829\n",
            "Epoch 76/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.1286 - mae: 13.1286\n",
            "Epoch 77/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.3276 - mae: 18.3276\n",
            "Epoch 78/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.0151 - mae: 9.0151\n",
            "Epoch 79/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 29.2670 - mae: 29.2670\n",
            "Epoch 80/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 31.6991 - mae: 31.6991\n",
            "Epoch 81/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.9313 - mae: 13.9313\n",
            "Epoch 82/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.6949 - mae: 15.6949\n",
            "Epoch 83/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.1858 - mae: 11.1858\n",
            "Epoch 84/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.3165 - mae: 7.3165\n",
            "Epoch 85/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.1247 - mae: 15.1247\n",
            "Epoch 86/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 15.4154 - mae: 15.4154\n",
            "Epoch 87/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 19.3599 - mae: 19.3599\n",
            "Epoch 88/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.1743 - mae: 29.1743\n",
            "Epoch 89/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.0967 - mae: 10.0967\n",
            "Epoch 90/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 20.9884 - mae: 20.9884\n",
            "Epoch 91/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4589 - mae: 10.4589\n",
            "Epoch 92/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.9153 - mae: 17.9153\n",
            "Epoch 93/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 6.5722 - mae: 6.5722\n",
            "Epoch 94/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 11.1121 - mae: 11.1121\n",
            "Epoch 95/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 24.2696 - mae: 24.2696\n",
            "Epoch 96/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.7170 - mae: 10.7170\n",
            "Epoch 97/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.2950 - mae: 15.2950\n",
            "Epoch 98/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2424 - mae: 8.2424\n",
            "Epoch 99/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.4659 - mae: 16.4659\n",
            "Epoch 100/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.2617 - mae: 14.2617\n",
            "Epoch 101/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 15.2383 - mae: 15.2383\n",
            "Epoch 102/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.7465 - mae: 10.7465\n",
            "Epoch 103/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.1772 - mae: 9.1772\n",
            "Epoch 104/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.7473 - mae: 23.7473\n",
            "Epoch 105/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.8294 - mae: 10.8294\n",
            "Epoch 106/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.2837 - mae: 11.2837\n",
            "Epoch 107/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 21.2974 - mae: 21.2974\n",
            "Epoch 108/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 6.3310 - mae: 6.3310\n",
            "Epoch 109/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.6146 - mae: 10.6146\n",
            "Epoch 110/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.5534 - mae: 10.5534\n",
            "Epoch 111/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 16.7255 - mae: 16.7255\n",
            "Epoch 112/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.5318 - mae: 9.5318\n",
            "Epoch 113/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 17.6015 - mae: 17.6015\n",
            "Epoch 114/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.9734 - mae: 16.9734\n",
            "Epoch 115/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.0678 - mae: 11.0678\n",
            "Epoch 116/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.0208 - mae: 23.0208\n",
            "Epoch 117/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.5451 - mae: 9.5451\n",
            "Epoch 118/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.6015 - mae: 10.6015\n",
            "Epoch 119/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.0432 - mae: 8.0432\n",
            "Epoch 120/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.2934 - mae: 29.2934\n",
            "Epoch 121/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.0807 - mae: 8.0807\n",
            "Epoch 122/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 27.9776 - mae: 27.9776\n",
            "Epoch 123/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 32.4908 - mae: 32.4908\n",
            "Epoch 124/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 19.3273 - mae: 19.3273\n",
            "Epoch 125/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.5183 - mae: 9.5183\n",
            "Epoch 126/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.5928 - mae: 9.5928\n",
            "Epoch 127/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.6936 - mae: 12.6936\n",
            "Epoch 128/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 12.8008 - mae: 12.8008\n",
            "Epoch 129/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.9182 - mae: 13.9182\n",
            "Epoch 130/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.2348 - mae: 10.2348\n",
            "Epoch 131/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 21.6632 - mae: 21.6632\n",
            "Epoch 132/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.2769 - mae: 8.2769\n",
            "Epoch 133/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.0581 - mae: 9.0581\n",
            "Epoch 134/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.8729 - mae: 16.8729\n",
            "Epoch 135/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.6488 - mae: 10.6488\n",
            "Epoch 136/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.4818 - mae: 18.4818\n",
            "Epoch 137/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.4573 - mae: 23.4573\n",
            "Epoch 138/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2678 - mae: 9.2678\n",
            "Epoch 139/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.0065 - mae: 9.0065\n",
            "Epoch 140/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.9290 - mae: 16.9290\n",
            "Epoch 141/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.3246 - mae: 8.3246\n",
            "Epoch 142/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 33.9643 - mae: 33.9643\n",
            "Epoch 143/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.0426 - mae: 23.0426\n",
            "Epoch 144/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3457 - mae: 11.3457\n",
            "Epoch 145/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 25.0552 - mae: 25.0552\n",
            "Epoch 146/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.1298 - mae: 11.1298\n",
            "Epoch 147/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.1057 - mae: 14.1057\n",
            "Epoch 148/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.9605 - mae: 16.9605\n",
            "Epoch 149/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2357 - mae: 9.2357\n",
            "Epoch 150/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 7.9349 - mae: 7.9349\n",
            "Epoch 151/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.4477 - mae: 16.4477\n",
            "Epoch 152/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.7902 - mae: 9.7902\n",
            "Epoch 153/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 26.7510 - mae: 26.7510\n",
            "Epoch 154/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.9444 - mae: 11.9444\n",
            "Epoch 155/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.3064 - mae: 15.3064\n",
            "Epoch 156/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.6946 - mae: 16.6946\n",
            "Epoch 157/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.1952 - mae: 19.1952\n",
            "Epoch 158/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.2366 - mae: 8.2366\n",
            "Epoch 159/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.9971 - mae: 7.9971\n",
            "Epoch 160/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 21.0029 - mae: 21.0029\n",
            "Epoch 161/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 23.6912 - mae: 23.6912\n",
            "Epoch 162/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.6280 - mae: 18.6280\n",
            "Epoch 163/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.5843 - mae: 17.5843\n",
            "Epoch 164/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.0288 - mae: 11.0288\n",
            "Epoch 165/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.6324 - mae: 9.6324\n",
            "Epoch 166/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 21.6074 - mae: 21.6074\n",
            "Epoch 167/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 26.2513 - mae: 26.2513\n",
            "Epoch 168/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.8701 - mae: 9.8701\n",
            "Epoch 169/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 22.5133 - mae: 22.5133\n",
            "Epoch 170/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.1706 - mae: 10.1706\n",
            "Epoch 171/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 18.0742 - mae: 18.0742\n",
            "Epoch 172/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 28.7735 - mae: 28.7735\n",
            "Epoch 173/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.4848 - mae: 16.4848\n",
            "Epoch 174/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 11.2286 - mae: 11.2286\n",
            "Epoch 175/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 27.6023 - mae: 27.6023\n",
            "Epoch 176/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.2709 - mae: 8.2709\n",
            "Epoch 177/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.2665 - mae: 9.2665\n",
            "Epoch 178/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 18.0801 - mae: 18.0801\n",
            "Epoch 179/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.6352 - mae: 10.6352\n",
            "Epoch 180/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.9094 - mae: 7.9094\n",
            "Epoch 181/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 17.3373 - mae: 17.3373\n",
            "Epoch 182/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.0011 - mae: 11.0011\n",
            "Epoch 183/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 11.6762 - mae: 11.6762\n",
            "Epoch 184/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 30.2799 - mae: 30.2799\n",
            "Epoch 185/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 8.1946 - mae: 8.1946\n",
            "Epoch 186/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 18.6861 - mae: 18.6861\n",
            "Epoch 187/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.7814 - mae: 8.7814\n",
            "Epoch 188/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.6678 - mae: 23.6678\n",
            "Epoch 189/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.4110 - mae: 9.4110\n",
            "Epoch 190/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.1721 - mae: 17.1721\n",
            "Epoch 191/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.5727 - mae: 8.5727\n",
            "Epoch 192/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.1662 - mae: 15.1662\n",
            "Epoch 193/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.9984 - mae: 29.9984\n",
            "Epoch 194/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.6085 - mae: 9.6085\n",
            "Epoch 195/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.9996 - mae: 11.9996\n",
            "Epoch 196/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.3435 - mae: 23.3435\n",
            "Epoch 197/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.1578 - mae: 18.1578\n",
            "Epoch 198/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.8348 - mae: 10.8348\n",
            "Epoch 199/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 22.4461 - mae: 22.4461\n",
            "Epoch 200/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.0942 - mae: 17.0942\n",
            "Epoch 201/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.3949 - mae: 7.3949\n",
            "Epoch 202/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 26.0937 - mae: 26.0937\n",
            "Epoch 203/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.3877 - mae: 13.3877\n",
            "Epoch 204/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.4446 - mae: 18.4446\n",
            "Epoch 205/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.2438 - mae: 9.2438\n",
            "Epoch 206/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.3339 - mae: 10.3339\n",
            "Epoch 207/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 20.6810 - mae: 20.6810\n",
            "Epoch 208/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.1771 - mae: 16.1771\n",
            "Epoch 209/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.0812 - mae: 14.0812\n",
            "Epoch 210/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 18.9659 - mae: 18.9659\n",
            "Epoch 211/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 10.2158 - mae: 10.2158\n",
            "Epoch 212/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 19.5664 - mae: 19.5664\n",
            "Epoch 213/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 14.7254 - mae: 14.7254\n",
            "Epoch 214/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.3897 - mae: 14.3897\n",
            "Epoch 215/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 22.8832 - mae: 22.8832\n",
            "Epoch 216/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 14.3715 - mae: 14.3715\n",
            "Epoch 217/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 9.1504 - mae: 9.1504\n",
            "Epoch 218/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.9336 - mae: 11.9336\n",
            "Epoch 219/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 6.7748 - mae: 6.7748\n",
            "Epoch 220/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.1318 - mae: 7.1318\n",
            "Epoch 221/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 38.3824 - mae: 38.3824\n",
            "Epoch 222/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 38.3848 - mae: 38.3848\n",
            "Epoch 223/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 5.5119 - mae: 5.5119\n",
            "Epoch 224/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.0322 - mae: 15.0322\n",
            "Epoch 225/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 17.0960 - mae: 17.0960\n",
            "Epoch 226/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.2786 - mae: 16.2786\n",
            "Epoch 227/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 16.6253 - mae: 16.6253\n",
            "Epoch 228/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.0301 - mae: 10.0301\n",
            "Epoch 229/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.9509 - mae: 17.9509\n",
            "Epoch 230/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.5546 - mae: 15.5546\n",
            "Epoch 231/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 21.0481 - mae: 21.0481\n",
            "Epoch 232/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 25.6419 - mae: 25.6419\n",
            "Epoch 233/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 16.6140 - mae: 16.6140\n",
            "Epoch 234/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.3947 - mae: 7.3947\n",
            "Epoch 235/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.2520 - mae: 17.2520\n",
            "Epoch 236/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.2644 - mae: 7.2644\n",
            "Epoch 237/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.3754 - mae: 9.3754\n",
            "Epoch 238/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2111 - mae: 8.2111\n",
            "Epoch 239/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 17.2595 - mae: 17.2595\n",
            "Epoch 240/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.9931 - mae: 8.9931\n",
            "Epoch 241/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.3310 - mae: 13.3310\n",
            "Epoch 242/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.8934 - mae: 8.8934\n",
            "Epoch 243/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 18.9980 - mae: 18.9980\n",
            "Epoch 244/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.0270 - mae: 14.0270\n",
            "Epoch 245/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.6579 - mae: 14.6579\n",
            "Epoch 246/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 15.7949 - mae: 15.7949\n",
            "Epoch 247/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.6670 - mae: 17.6670\n",
            "Epoch 248/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.2255 - mae: 13.2255\n",
            "Epoch 249/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.5120 - mae: 14.5120\n",
            "Epoch 250/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 23.2205 - mae: 23.2205\n",
            "Epoch 251/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.3563 - mae: 9.3563\n",
            "Epoch 252/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 36.7027 - mae: 36.7027\n",
            "Epoch 253/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 21.8702 - mae: 21.8702\n",
            "Epoch 254/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 7.3342 - mae: 7.3342\n",
            "Epoch 255/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 24.7324 - mae: 24.7324\n",
            "Epoch 256/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 12.4683 - mae: 12.4683\n",
            "Epoch 257/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.6237 - mae: 10.6237\n",
            "Epoch 258/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 14.2401 - mae: 14.2401\n",
            "Epoch 259/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.2911 - mae: 11.2911\n",
            "Epoch 260/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 31.6518 - mae: 31.6518\n",
            "Epoch 261/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.2234 - mae: 11.2234\n",
            "Epoch 262/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.0738 - mae: 10.0738\n",
            "Epoch 263/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.9785 - mae: 8.9785\n",
            "Epoch 264/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 21.5218 - mae: 21.5218\n",
            "Epoch 265/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.4489 - mae: 11.4489\n",
            "Epoch 266/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 13.2883 - mae: 13.2883\n",
            "Epoch 267/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.0868 - mae: 11.0868\n",
            "Epoch 268/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.1521 - mae: 19.1521\n",
            "Epoch 269/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 40.7220 - mae: 40.7220\n",
            "Epoch 270/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.9472 - mae: 12.9472\n",
            "Epoch 271/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.7670 - mae: 14.7670\n",
            "Epoch 272/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 28.5086 - mae: 28.5086\n",
            "Epoch 273/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.3896 - mae: 7.3896\n",
            "Epoch 274/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 6.3715 - mae: 6.3715\n",
            "Epoch 275/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 36.9084 - mae: 36.9084\n",
            "Epoch 276/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.3074 - mae: 8.3074\n",
            "Epoch 277/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 27.7263 - mae: 27.7263\n",
            "Epoch 278/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.6850 - mae: 10.6850\n",
            "Epoch 279/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 16.1133 - mae: 16.1133\n",
            "Epoch 280/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 21.2362 - mae: 21.2362\n",
            "Epoch 281/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 23.7909 - mae: 23.7909\n",
            "Epoch 282/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2787 - mae: 8.2787\n",
            "Epoch 283/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.4584 - mae: 8.4584\n",
            "Epoch 284/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 26.6233 - mae: 26.6233\n",
            "Epoch 285/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.2295 - mae: 14.2295\n",
            "Epoch 286/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.2749 - mae: 5.2749\n",
            "Epoch 287/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 20.8216 - mae: 20.8216\n",
            "Epoch 288/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 27.6077 - mae: 27.6077\n",
            "Epoch 289/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.4399 - mae: 10.4399\n",
            "Epoch 290/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.3118 - mae: 16.3118\n",
            "Epoch 291/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.5002 - mae: 16.5002\n",
            "Epoch 292/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.4403 - mae: 7.4403\n",
            "Epoch 293/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.5514 - mae: 16.5514\n",
            "Epoch 294/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 25.3363 - mae: 25.3363\n",
            "Epoch 295/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.5777 - mae: 14.5777\n",
            "Epoch 296/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 4.6817 - mae: 4.6817\n",
            "Epoch 297/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.2923 - mae: 7.2923\n",
            "Epoch 298/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 19.6082 - mae: 19.6082\n",
            "Epoch 299/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.6275 - mae: 6.6275\n",
            "Epoch 300/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 22.0541 - mae: 22.0541\n",
            "Epoch 301/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.1444 - mae: 9.1444\n",
            "Epoch 302/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 11.1960 - mae: 11.1960\n",
            "Epoch 303/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.3465 - mae: 9.3465\n",
            "Epoch 304/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.7073 - mae: 23.7073\n",
            "Epoch 305/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.9304 - mae: 7.9304\n",
            "Epoch 306/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.7560 - mae: 23.7560\n",
            "Epoch 307/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.8904 - mae: 5.8904\n",
            "Epoch 308/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 21.3255 - mae: 21.3255\n",
            "Epoch 309/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 18.4083 - mae: 18.4083\n",
            "Epoch 310/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 4.5052 - mae: 4.5052\n",
            "Epoch 311/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 6.5155 - mae: 6.5155\n",
            "Epoch 312/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 35.4446 - mae: 35.4446\n",
            "Epoch 313/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 28.9225 - mae: 28.9225\n",
            "Epoch 314/500\n",
            "2/2 [==============================] - 0s 5ms/step - loss: 11.1961 - mae: 11.1961\n",
            "Epoch 315/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.5542 - mae: 23.5542\n",
            "Epoch 316/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.4229 - mae: 14.4229\n",
            "Epoch 317/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 19.9704 - mae: 19.9704\n",
            "Epoch 318/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.0022 - mae: 8.0022\n",
            "Epoch 319/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 18.1053 - mae: 18.1053\n",
            "Epoch 320/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.6392 - mae: 10.6392\n",
            "Epoch 321/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.0277 - mae: 7.0277\n",
            "Epoch 322/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.6763 - mae: 8.6763\n",
            "Epoch 323/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.3415 - mae: 18.3415\n",
            "Epoch 324/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 6.3087 - mae: 6.3087\n",
            "Epoch 325/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 14.6078 - mae: 14.6078\n",
            "Epoch 326/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 6.9910 - mae: 6.9910\n",
            "Epoch 327/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.6168 - mae: 17.6168\n",
            "Epoch 328/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.3761 - mae: 14.3761\n",
            "Epoch 329/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.6896 - mae: 17.6896\n",
            "Epoch 330/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.7835 - mae: 6.7835\n",
            "Epoch 331/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 19.7339 - mae: 19.7339\n",
            "Epoch 332/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.5491 - mae: 10.5491\n",
            "Epoch 333/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 16.3341 - mae: 16.3341\n",
            "Epoch 334/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.7210 - mae: 9.7210\n",
            "Epoch 335/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 13.0579 - mae: 13.0579\n",
            "Epoch 336/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 32.4942 - mae: 32.4942\n",
            "Epoch 337/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.0392 - mae: 11.0392\n",
            "Epoch 338/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 19.9548 - mae: 19.9548\n",
            "Epoch 339/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 34.3127 - mae: 34.3127\n",
            "Epoch 340/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.7224 - mae: 8.7224\n",
            "Epoch 341/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 21.9581 - mae: 21.9581\n",
            "Epoch 342/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 13.8445 - mae: 13.8445\n",
            "Epoch 343/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 11.6902 - mae: 11.6902\n",
            "Epoch 344/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.6927 - mae: 10.6927\n",
            "Epoch 345/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 30.9047 - mae: 30.9047\n",
            "Epoch 346/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.6409 - mae: 10.6409\n",
            "Epoch 347/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 25.5635 - mae: 25.5635\n",
            "Epoch 348/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.2869 - mae: 13.2869\n",
            "Epoch 349/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.0354 - mae: 13.0354\n",
            "Epoch 350/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.4008 - mae: 15.4008\n",
            "Epoch 351/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 32.8738 - mae: 32.8738\n",
            "Epoch 352/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.0835 - mae: 14.0835\n",
            "Epoch 353/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.7779 - mae: 17.7779\n",
            "Epoch 354/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3423 - mae: 11.3423\n",
            "Epoch 355/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 26.7688 - mae: 26.7688\n",
            "Epoch 356/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.1525 - mae: 10.1525\n",
            "Epoch 357/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.7123 - mae: 14.7123\n",
            "Epoch 358/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 14.5868 - mae: 14.5868\n",
            "Epoch 359/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.3507 - mae: 12.3507\n",
            "Epoch 360/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 20.3633 - mae: 20.3633\n",
            "Epoch 361/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 10.8795 - mae: 10.8795\n",
            "Epoch 362/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.8191 - mae: 6.8191\n",
            "Epoch 363/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 23.7855 - mae: 23.7855\n",
            "Epoch 364/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.4773 - mae: 29.4773\n",
            "Epoch 365/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.2764 - mae: 8.2764\n",
            "Epoch 366/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.1103 - mae: 6.1103\n",
            "Epoch 367/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 34.5598 - mae: 34.5598\n",
            "Epoch 368/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.3526 - mae: 7.3526\n",
            "Epoch 369/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.7014 - mae: 8.7014\n",
            "Epoch 370/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.2519 - mae: 14.2519\n",
            "Epoch 371/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.8540 - mae: 6.8540\n",
            "Epoch 372/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.5604 - mae: 6.5604\n",
            "Epoch 373/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 24.1993 - mae: 24.1993\n",
            "Epoch 374/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.2210 - mae: 10.2210\n",
            "Epoch 375/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.8456 - mae: 12.8456\n",
            "Epoch 376/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.7178 - mae: 14.7178\n",
            "Epoch 377/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.6495 - mae: 14.6495\n",
            "Epoch 378/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.1752 - mae: 16.1752\n",
            "Epoch 379/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 20.5996 - mae: 20.5996\n",
            "Epoch 380/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 34.2181 - mae: 34.2181\n",
            "Epoch 381/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 8.8638 - mae: 8.8638\n",
            "Epoch 382/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 9.6235 - mae: 9.6235\n",
            "Epoch 383/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.7712 - mae: 5.7712\n",
            "Epoch 384/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.7087 - mae: 8.7087\n",
            "Epoch 385/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 4.9657 - mae: 4.9657\n",
            "Epoch 386/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 25.1170 - mae: 25.1170\n",
            "Epoch 387/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 15.1353 - mae: 15.1353\n",
            "Epoch 388/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 6.6467 - mae: 6.6467\n",
            "Epoch 389/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 17.5620 - mae: 17.5620\n",
            "Epoch 390/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.8474 - mae: 23.8474\n",
            "Epoch 391/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 16.5450 - mae: 16.5450\n",
            "Epoch 392/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.3678 - mae: 8.3678\n",
            "Epoch 393/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 18.1358 - mae: 18.1358\n",
            "Epoch 394/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 14.1971 - mae: 14.1971\n",
            "Epoch 395/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 28.6632 - mae: 28.6632\n",
            "Epoch 396/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 8.2331 - mae: 8.2331\n",
            "Epoch 397/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 10.5301 - mae: 10.5301\n",
            "Epoch 398/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.4039 - mae: 7.4039\n",
            "Epoch 399/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 15.6112 - mae: 15.6112\n",
            "Epoch 400/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 6.7504 - mae: 6.7504\n",
            "Epoch 401/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.0176 - mae: 8.0176\n",
            "Epoch 402/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 16.4587 - mae: 16.4587\n",
            "Epoch 403/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.3651 - mae: 12.3651\n",
            "Epoch 404/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 22.9409 - mae: 22.9409\n",
            "Epoch 405/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 18.0841 - mae: 18.0841\n",
            "Epoch 406/500\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 7.0684 - mae: 7.0684\n",
            "Epoch 407/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.6386 - mae: 12.6386\n",
            "Epoch 408/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 5.6819 - mae: 5.6819\n",
            "Epoch 409/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 31.1786 - mae: 31.1786\n",
            "Epoch 410/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 9.2730 - mae: 9.2730\n",
            "Epoch 411/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.8854 - mae: 14.8854\n",
            "Epoch 412/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 21.7679 - mae: 21.7679\n",
            "Epoch 413/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 12.5743 - mae: 12.5743\n",
            "Epoch 414/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 6.0794 - mae: 6.0794\n",
            "Epoch 415/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 13.2388 - mae: 13.2388\n",
            "Epoch 416/500\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 27.4611 - mae: 27.4611\n",
            "Epoch 417/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.5558 - mae: 10.5558\n",
            "Epoch 418/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 12.8502 - mae: 12.8502\n",
            "Epoch 419/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 15.8979 - mae: 15.8979\n",
            "Epoch 420/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 24.7882 - mae: 24.7882\n",
            "Epoch 421/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 17.2852 - mae: 17.2852\n",
            "Epoch 422/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 7.8194 - mae: 7.8194\n",
            "Epoch 423/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 25.4122 - mae: 25.4122\n",
            "Epoch 424/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 15.0304 - mae: 15.0304\n",
            "Epoch 425/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 7.1304 - mae: 7.1304\n",
            "Epoch 426/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 20.2448 - mae: 20.2448\n",
            "Epoch 427/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 6.2939 - mae: 6.2939\n",
            "Epoch 428/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.9396 - mae: 12.9396\n",
            "Epoch 429/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 10.7375 - mae: 10.7375\n",
            "Epoch 430/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 11.3486 - mae: 11.3486\n",
            "Epoch 431/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 10.5659 - mae: 10.5659\n",
            "Epoch 432/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 11.3970 - mae: 11.3970\n",
            "Epoch 433/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.3850 - mae: 11.3850\n",
            "Epoch 434/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 30.4177 - mae: 30.4177\n",
            "Epoch 435/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.5036 - mae: 10.5036\n",
            "Epoch 436/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 28.9014 - mae: 28.9014\n",
            "Epoch 437/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 8.6075 - mae: 8.6075\n",
            "Epoch 438/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 12.7373 - mae: 12.7373\n",
            "Epoch 439/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 33.6882 - mae: 33.6882\n",
            "Epoch 440/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 15.1078 - mae: 15.1078\n",
            "Epoch 441/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.4894 - mae: 17.4894\n",
            "Epoch 442/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 22.3290 - mae: 22.3290\n",
            "Epoch 443/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 23.5265 - mae: 23.5265\n",
            "Epoch 444/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 10.9499 - mae: 10.9499\n",
            "Epoch 445/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 14.9282 - mae: 14.9282\n",
            "Epoch 446/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 18.0143 - mae: 18.0143\n",
            "Epoch 447/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 5.4175 - mae: 5.4175\n",
            "Epoch 448/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.0649 - mae: 10.0649\n",
            "Epoch 449/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.0152 - mae: 14.0152\n",
            "Epoch 450/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 16.7867 - mae: 16.7867\n",
            "Epoch 451/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 14.2944 - mae: 14.2944\n",
            "Epoch 452/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 30.6302 - mae: 30.6302\n",
            "Epoch 453/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 7.6477 - mae: 7.6477\n",
            "Epoch 454/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 28.1540 - mae: 28.1540\n",
            "Epoch 455/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 7.9732 - mae: 7.9732\n",
            "Epoch 456/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 10.3972 - mae: 10.3972\n",
            "Epoch 457/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.0323 - mae: 15.0323\n",
            "Epoch 458/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 16.5658 - mae: 16.5658\n",
            "Epoch 459/500\n",
            "2/2 [==============================] - 0s 6ms/step - loss: 26.8667 - mae: 26.8667\n",
            "Epoch 460/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.4613 - mae: 12.4613\n",
            "Epoch 461/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.4836 - mae: 12.4836\n",
            "Epoch 462/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 13.3210 - mae: 13.3210\n",
            "Epoch 463/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 29.5603 - mae: 29.5603\n",
            "Epoch 464/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 3.4566 - mae: 3.4566\n",
            "Epoch 465/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.2194 - mae: 15.2194\n",
            "Epoch 466/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 20.8425 - mae: 20.8425\n",
            "Epoch 467/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 30.4862 - mae: 30.4862\n",
            "Epoch 468/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 11.0427 - mae: 11.0427\n",
            "Epoch 469/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 12.8198 - mae: 12.8198\n",
            "Epoch 470/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 3.2258 - mae: 3.2258\n",
            "Epoch 471/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 16.6991 - mae: 16.6991\n",
            "Epoch 472/500\n",
            "2/2 [==============================] - 0s 12ms/step - loss: 13.3916 - mae: 13.3916\n",
            "Epoch 473/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 15.2781 - mae: 15.2781\n",
            "Epoch 474/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.7453 - mae: 11.7453\n",
            "Epoch 475/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 16.4156 - mae: 16.4156\n",
            "Epoch 476/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 13.8775 - mae: 13.8775\n",
            "Epoch 477/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 30.6744 - mae: 30.6744\n",
            "Epoch 478/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 8.5918 - mae: 8.5918\n",
            "Epoch 479/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 10.7364 - mae: 10.7364\n",
            "Epoch 480/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 17.9045 - mae: 17.9045\n",
            "Epoch 481/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 15.8082 - mae: 15.8082\n",
            "Epoch 482/500\n",
            "2/2 [==============================] - 0s 11ms/step - loss: 21.3081 - mae: 21.3081\n",
            "Epoch 483/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 25.3807 - mae: 25.3807\n",
            "Epoch 484/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.9720 - mae: 23.9720\n",
            "Epoch 485/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 5.7608 - mae: 5.7608\n",
            "Epoch 486/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 19.9963 - mae: 19.9963\n",
            "Epoch 487/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 14.0412 - mae: 14.0412\n",
            "Epoch 488/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 30.6104 - mae: 30.6104\n",
            "Epoch 489/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 11.9403 - mae: 11.9403\n",
            "Epoch 490/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 12.7339 - mae: 12.7339\n",
            "Epoch 491/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.6144 - mae: 23.6144\n",
            "Epoch 492/500\n",
            "2/2 [==============================] - 0s 10ms/step - loss: 20.5371 - mae: 20.5371\n",
            "Epoch 493/500\n",
            "2/2 [==============================] - 0s 9ms/step - loss: 4.9829 - mae: 4.9829\n",
            "Epoch 494/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 12.7955 - mae: 12.7955\n",
            "Epoch 495/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 13.3751 - mae: 13.3751\n",
            "Epoch 496/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 12.6708 - mae: 12.6708\n",
            "Epoch 497/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 17.6143 - mae: 17.6143\n",
            "Epoch 498/500\n",
            "2/2 [==============================] - 0s 8ms/step - loss: 23.5694 - mae: 23.5694\n",
            "Epoch 499/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 9.3806 - mae: 9.3806\n",
            "Epoch 500/500\n",
            "2/2 [==============================] - 0s 7ms/step - loss: 14.6271 - mae: 14.6271\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba28835c580>"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make and plot some predictions\n",
        "y_preds_3 = model_3.predict(X_test)\n",
        "plot_predictions(predictions=y_preds_3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 616
        },
        "id": "la-Me_T9iWRE",
        "outputId": "29372ff8-5a53-448d-d70b-cf35160c4a9e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 71ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x700 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0EAAAJGCAYAAACdj47VAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABjt0lEQVR4nO3dfXyT9bnH8W8aaAWhrSDQ0IQHRYGJj7hxYEbLkQnqtBA7POAUnMNtglLRTTnH+bTt4Hxaq3NTpxbPpk7s7jHmNhygaFRAxmTMrSCwYmsIsKm0iPKU3ueP22RN05K0zXM+79crr3L/7utOf8li16u/331dNtM0TQEAAABAjshL9QQAAAAAIJlIggAAAADkFJIgAAAAADmFJAgAAABATiEJAgAAAJBTSIIAAAAA5BSSIAAAAAA5pUeqJ9BdLS0t2rlzp/r27SubzZbq6QAAAABIEdM0tW/fPg0ePFh5eR2v92R8ErRz5065XK5UTwMAAABAmmhsbJTT6ezwfMYnQX379pVkvdDCwsIUzwYAAABAqjQ3N8vlcoVyhI5kfBIU3AJXWFhIEgQAAAAg6m0yFEYAAAAAkFNIggAAAADkFJIgAAAAADkl4+8JilUgENDhw4dTPQ2ksZ49e8put6d6GgAAAEiwrE+CTNPUrl27tHfv3lRPBRmguLhYJSUl9JwCAADIYlmfBAUToIEDB6p37978cot2maapTz75RHv27JEkORyOFM8IAAAAiZLVSVAgEAglQP3790/1dJDmevXqJUnas2ePBg4cyNY4AACALJXVhRGC9wD17t07xTNBpgh+Vrh/DAAAIHtldRIUxBY4xIrPCgAAQPbLiSQIAAAAAIJIgnLEsGHDVFVVFXP86tWrZbPZUlJVb/HixSouLk769wUAAEBuIAlKMzab7aiPO++8s0vPu379el177bUxx0+YMEF+v19FRUVd+n7J1tkkDwAAALkrq6vDxUsgIHm9kt8vORyS2y0lqnCY3+8P/fv555/X7bffri1btoTG+vTpE/q3aZoKBALq0SP6/4wDBgzo1Dzy8/NVUlLSqWsAAACATMBKUBSGIQ0bJk2cKM2caX0dNswaT4SSkpLQo6ioSDabLXS8efNm9e3bV3/4wx80duxYFRQU6PXXX9f27dtVXl6uQYMGqU+fPvr85z+vlStXhj1v25USm82mJ554QtOmTVPv3r110kknadmyZaHzbbfDBbeovfTSSxo9erT69OmjKVOmhCVtR44c0Q033KDi4mL1799ft9xyi2bNmqWpU6ce9TUvXrxYQ4YMUe/evTVt2jR98MEHYeejvb6ysjK99957uvHGG0MrZpL0wQcfaMaMGSotLVXv3r116qmn6rnnnuvM/xwAAADIQiRBR2EYUkWF9P774eM+nzWeqEQomltvvVX33HOP6urqdNppp+njjz/WRRddpFWrVuntt9/WlClTdMkll6ihoeGoz3PXXXdp+vTp2rRpky666CJdccUV+vDDDzuM/+STT3T//ffr5z//uV577TU1NDTo5ptvDp3/4Q9/qGeeeUY1NTV644031NzcrKVLlx51DuvWrdM111yjefPmaePGjZo4caK+//3vh8VEe32GYcjpdOruu++W3+8PJWYHDhzQ2LFj9bvf/U7vvPOOrr32Wl155ZV66623jjonAAAAZDkzwzU1NZmSzKampohzn376qfn3v//d/PTTTzv9vEeOmKbTaZpS+w+bzTRdLisuUWpqasyioqLQ8SuvvGJKMpcuXRr12lNOOcV8+OGHQ8dDhw41f/SjH4WOJZm33XZb6Pjjjz82JZl/+MMfwr7XRx99FJqLJHPbtm2hax555BFz0KBBoeNBgwaZ9913X+j4yJEj5pAhQ8zy8vIO5zljxgzzoosuChu7/PLLw153V15fRy6++GLzpptu6vB8dz4zAAAASK2j5QatsRLUAa83cgWoNdOUGhutuGQ7++yzw44//vhj3XzzzRo9erSKi4vVp08f1dXVRV0JOu2000L/PvbYY1VYWKg9e/Z0GN+7d2+deOKJoWOHwxGKb2pq0u7du/WFL3whdN5ut2vs2LFHnUNdXZ3GjRsXNjZ+/Pi4vL5AIKDvfe97OvXUU9WvXz/16dNHL730UtTrAAAAkN0ojNCBVre6xCUuno499tiw45tvvlkrVqzQ/fffrxEjRqhXr16qqKjQoUOHjvo8PXv2DDu22WxqaWnpVLxpmp2cfed19fXdd999qq6uVlVVlU499VQde+yxqqysjHodAAAAYhNoCcjb4JV/n1+Ovg65h7hlz0tQBbE4IgnqgMMR37hEeuONNzR79mxNmzZNkrVysmPHjqTOoaioSIMGDdL69et17rnnSrJWYv785z/rjDPO6PC60aNHa926dWFja9euDTuO5fXl5+crEAhEXFdeXq6vfvWrkqSWlha9++67+tznPteVlwgAAIBWjDpD85fP1/vN/94+5Sx0qnpKtTyjPSmcWXRsh+uA2y05ndJnhcYi2GySy2XFpdpJJ50kwzC0ceNG/eUvf9HMmTOPuqKTKNdff70WLVqk3/zmN9qyZYvmz5+vjz76KFStrT033HCDli9frvvvv19bt27Vj3/8Yy1fvjwsJpbXN2zYML322mvy+Xz617/+FbpuxYoVevPNN1VXV6dvfOMb2r17d/xfOAAAQI4x6gxVLKkIS4AkydfsU8WSChl1KaogFiOSoA7Y7VJ1tfXvtr/DB4+rqhLXL6gzHnzwQR133HGaMGGCLrnkEk2ePFlnnXVW0udxyy23aMaMGbrqqqs0fvx49enTR5MnT9YxxxzT4TX/8R//oZ/97Geqrq7W6aefrj/+8Y+67bbbwmJieX133323duzYoRNPPDHUE+m2227TWWedpcmTJ6usrEwlJSVRy3UDAADg6AItAc1fPl+mIm+LCI5VLq9UoCUQcT5d2Mxk3NSRQM3NzSoqKlJTU5MKCwvDzh04cED19fUaPnz4UX8RPxrDkObPDy+S4HJZCZAnvVf5Uq6lpUWjR4/W9OnT9b3vfS/V04lJPD4zAAAA2Wz1jtWa+PTEqHGvzHpFZcPKEj+hVo6WG7TGPUFReDxSeblVBc7vt+4BcrvTYwUo3bz33nv64x//qPPOO08HDx7Uj3/8Y9XX12vmzJmpnhoAAADixL8vtspgscalAklQDOx2qaws1bNIf3l5eVq8eLFuvvlmmaapMWPGaOXKlRo9enSqpwYAAIA4cfSNrTJYrHGpQBKEuHG5XHrjjTdSPQ0AAAAkkHuIW85Cp3zNvnbvC7LJJmehU+4haVBBrAMURgAAAAAQM3ueXdVTrApiNoVXEAseV02pSut+QSRBAAAAADrFM9qj2um1Ki0sDRt3FjpVO7027fsEsR0OAAAAyGGBloC8DV759/nl6OuQe4g7plUcz2iPykeWd+naVCMJAgAAAHKUUWdo/vL5YU1PnYVOVU+pjmk1x55nT3oZ7HhgOxwAAACQg4w6QxVLKsISIEnyNftUsaRCRp2RopklHkkQAAAAkGMCLQHNXz6/3epuwbHK5ZUKtASSPbWkIAnKcXfeeafOOOOMlHzv2bNna+rUqSn53gAAALnM2+CNWAFqzZSpxuZGeRu8SZxV8pAEpRmbzXbUx5133tmt5166dGnY2M0336xVq1Z1b9JJsmPHDtlsNm3cuDHVUwEAAMho/n3+uMZlmi4nQa+99pouueQSDR48uN1frk3T1O233y6Hw6FevXpp0qRJ2rp1a1jMhx9+qCuuuEKFhYUqLi7WNddco48//rirU0qYQEtAq3es1nN/fU6rd6xO6LKg3+8PPaqqqlRYWBg2dvPNN8f1+/Xp00f9+/eP63MCAAAgvTn6OuIal2m6nATt379fp59+uh555JF2z99777166KGH9Oijj2rdunU69thjNXnyZB04cCAUc8UVV+hvf/ubVqxYoRdffFGvvfaarr322q5OKSGMOkPDqodp4tMTNdOYqYlPT9Sw6mEJu1GspKQk9CgqKpLNZgsb++Uvf6nRo0frmGOO0ahRo/STn/wkdO2hQ4c0b948ORwOHXPMMRo6dKgWLVokSRo2bJgkadq0abLZbKHjttvhglvU7r//fjkcDvXv319z587V4cOHQzF+v18XX3yxevXqpeHDh+vZZ5/VsGHDVFVV1eHrCgQCWrBggYqLi9W/f3995zvfkWmG70Fdvny5zjnnnFDMl7/8ZW3fvj10fvjw4ZKkM888UzabTWVlZZKk9evX60tf+pKOP/54FRUV6bzzztOf//znzr71AAAAOcM9xC1noTOi2WmQTTa5Cl1yD3EneWbJ0eUk6MILL9T3v/99TZs2LeKcaZqqqqrSbbfdpvLycp122mn6v//7P+3cuTO0YlRXV6fly5friSee0Lhx43TOOefo4Ycf1i9/+Uvt3Lmzyy8ontKtYsYzzzyj22+/XT/4wQ9UV1en//3f/9V3v/tdPf3005Kkhx56SMuWLdOSJUu0ZcsWPfPMM6FkZ/369ZKkmpoa+f3+0HF7XnnlFW3fvl2vvPKKnn76aS1evFiLFy8Onb/qqqu0c+dOrV69Wr/61a/0+OOPa8+ePUed+wMPPKDFixfrqaee0uuvv64PP/xQv/71r8Ni9u/frwULFuhPf/qTVq1apby8PE2bNk0tLS2SpLfeekuStHLlSvn9fhmG9f7v27dPs2bN0uuvv661a9fqpJNO0kUXXaR9+/bF/uYCAADkEHueXdVTqiUpIhEKHldNqcqInj9dkZA+QfX19dq1a5cmTZoUGisqKtK4ceO0Zs0a/dd//ZfWrFmj4uJinX322aGYSZMmKS8vT+vWrWs3uZKkgwcP6uDBg6Hj5ubmRLyEqBUzbLKpcnmlykeWJ+3Dcccdd+iBBx6Qx2PVbB8+fLj+/ve/67HHHtOsWbPU0NCgk046Seecc45sNpuGDh0aunbAgAGSpOLiYpWUlBz1+xx33HH68Y9/LLvdrlGjRuniiy/WqlWrNGfOHG3evFkrV67U+vXrQ//bPfHEEzrppJOO+pxVVVVauHBhaO6PPvqoXnrppbCYyy67LOz4qaee0oABA/T3v/9dY8aMCb2G/v37h72G//zP/wy77vHHH1dxcbFeffVVffnLXz7qvAAAAHKVZ7RHtdNr2+0TVDWlKqY+QZkqIUnQrl27JEmDBg0KGx80aFDo3K5duzRw4MDwyfTooX79+oVi2rNo0SLdddddcZ5xpM5UzEhGg6j9+/dr+/btuuaaazRnzpzQ+JEjR1RUVCTJ2sr2pS99SSNHjtSUKVP05S9/WRdccEGnv9cpp5wiu/3fiZ3D4dBf//pXSdKWLVvUo0cPnXXWWaHzI0aM0HHHHdfh8zU1Ncnv92vcuHGhsR49eujss88O2xK3detW3X777Vq3bp3+9a9/hVaAGhoaNGbMmA6ff/fu3brtttu0evVq7dmzR4FAQJ988okaGho6/doBAAAyVaAlIG+DV/59fjn6OuQe4o76x3rPaI/KR5Z3+rpMl5AkKJEWLlyoBQsWhI6bm5vlcrni/n3SrWJGsGDEz372s7BkQlIoYTnrrLNUX1+vP/zhD1q5cqWmT5+uSZMmqba2tlPfq2fPnmHHNpstlJAk0iWXXKKhQ4fqZz/7mQYPHqyWlhaNGTNGhw4dOup1s2bN0gcffKDq6moNHTpUBQUFGj9+fNTrAAAAsoVRZ7S7olM9pTrqio49z56UP+qnk4SUyA5uVdq9e3fY+O7du0PnSkpKIu4jOXLkiD788MOjbtcqKChQYWFh2CMR0q1ixqBBgzR48GD94x//0IgRI8IewYIBklRYWKjLL79cP/vZz/T888/rV7/6lT788ENJVnITCHSvst3IkSN15MgRvf3226Gxbdu26aOPPurwmqKiIjkcDq1bty40duTIEW3YsCF0/MEHH2jLli267bbbdP7552v06NERz5mfny9JEa/hjTfe0A033KCLLrpIp5xyigoKCvSvf/2rW68TAAAgU6TbfeyZICFJ0PDhw1VSUhLWf6a5uVnr1q3T+PHjJUnjx4/X3r17w34Rfvnll9XS0hKx0pEK6Vgx46677tKiRYv00EMP6d1339Vf//pX1dTU6MEHH5QkPfjgg3ruuee0efNmvfvuu3rhhRdUUlKi4uJiSVaFuFWrVmnXrl1HTVqOZtSoUZo0aZKuvfZavfXWW3r77bd17bXXqlevXrLZ2n+vJGn+/Pm65557tHTpUm3evFnXXXed9u7dGzp/3HHHqX///nr88ce1bds2vfzyy2ErfpI0cOBA9erVS8uXL9fu3bvV1NQkSTrppJP085//XHV1dVq3bp2uuOIK9erVq0uvDwAAIJNEu49dkiqXVya0xUsm6nIS9PHHH2vjxo2hxpX19fXauHGjGhoaZLPZVFlZqe9///tatmyZ/vrXv+qqq67S4MGDNXXqVEnS6NGjNWXKFM2ZM0dvvfWW3njjDc2bN0//9V//pcGDB8fjtXVLOlbM+PrXv64nnnhCNTU1OvXUU3Xeeedp8eLFoZWgvn376t5779XZZ5+tz3/+89qxY4d+//vfKy/P+p/5gQce0IoVK+RyuXTmmWd2eR7/93//p0GDBuncc8/VtGnTNGfOHPXt21fHHHNMh9fcdNNNuvLKKzVr1iyNHz9effv2DSt+kZeXp1/+8pfasGGDxowZoxtvvFH33Xdf2HP06NFDDz30kB577DENHjxY5eXlkqQnn3xSH330kc466yxdeeWVuuGGGyLuNwMAAMhGnbmPHf9mM9s2a4nR6tWrNXHixIjxWbNmafHixTJNU3fccYcef/xx7d27V+ecc45+8pOf6OSTTw7Ffvjhh5o3b55++9vfKi8vT5dddpkeeugh9enTJ+Z5NDc3q6ioSE1NTRFb4w4cOKD6+noNHz78qL+gH017+ytdha6sr5jRGe+//75cLpdWrlyp888/P9XT6ZZ4fGYAAACS5bm/PqeZxsyocc96ntWMU2ckYUapdbTcoLUuJ0HpItFJkNS1ShvZ7OWXX9bHH3+sU089VX6/X9/5znfk8/n07rvvRhRVyDQkQQAAIJOs3rFaE5+OXJho65VZr+RE8YNYk6CMqw6XCrlYMeNoDh8+rP/+7//WP/7xD/Xt21cTJkzQM888k/EJEAAAQKYJ3sfua/a1e1+QTTY5C51JvY89E5AEodMmT56syZMnp3oaAAAAOS94H3vFkgrZZAtLhFJ1H3smSEh1OAAAAADJ4RntUe30WpUWloaNOwudqp1ey33s7WAlCAAAAEgTXb0X3TPao/KR5dzHHiOSIAAAACANtFeV2FnoVPWU6phWc7iPPXZshwMAAABSzKgzVLGkIqLnj6/Zp4olFTLqjBTNLDuRBAEAAAApFGgJaP7y+e1WdwuOVS6vVKAlkOypZS2SIAAAACCFvA3eiBWg1kyZamxulLfBm8RZZTeSoBw3e/ZsTZ06NXRcVlamysrKbj1nPJ4DAAAgV/j3+eMah+hIgtLU7NmzZbPZZLPZlJ+frxEjRujuu+/WkSNHEvp9DcPQ9773vZhiV69eLZvNpr1793b5OQAAAHKdo68jrnGIjupwsQgEJK9X8vslh0NyuyV74ssNTpkyRTU1NTp48KB+//vfa+7cuerZs6cWLlwYFnfo0CHl5+fH5Xv269cvLZ4DAAAgV7iHuOUsdMrX7Gv3viCbbHIWOuUe4k7B7LITK0HRGIY0bJg0caI0c6b1ddgwazzBCgoKVFJSoqFDh+pb3/qWJk2apGXLloW2sP3gBz/Q4MGDNXLkSElSY2Ojpk+fruLiYvXr10/l5eXasWNH6PkCgYAWLFig4uJi9e/fX9/5zndkmuH/obXdynbw4EHdcsstcrlcKigo0IgRI/Tkk09qx44dmjhxoiTpuOOOk81m0+zZs9t9jo8++khXXXWVjjvuOPXu3VsXXnihtm7dGjq/ePFiFRcX66WXXtLo0aPVp08fTZkyRX7/v5d8V69erS984Qs69thjVVxcrC9+8Yt677334vROAwAApI49z67qKdWSrISnteBx1ZQqev7EEUnQ0RiGVFEhvd/mRjWfzxpPQiLUWq9evXTo0CFJ0qpVq7RlyxatWLFCL774og4fPqzJkyerb9++8nq9euONN0LJRPCaBx54QIsXL9ZTTz2l119/XR9++KF+/etfH/V7XnXVVXruuef00EMPqa6uTo899pj69Okjl8ulX/3qV5KkLVu2yO/3q7q6ut3nmD17tv70pz9p2bJlWrNmjUzT1EUXXaTDhw+HYj755BPdf//9+vnPf67XXntNDQ0NuvnmmyVJR44c0dSpU3Xeeedp06ZNWrNmja699lrZbLZ2vx8AAECm8Yz2qHZ6rUoLS8PGnYVO1U6vjalPEGLHdriOBALS/PmSGbkkKdOUbDapslIqL0/41jjTNLVq1Sq99NJLuv766/XPf/5Txx57rJ544onQNrhf/OIXamlp0RNPPBFKDmpqalRcXKzVq1frggsuUFVVlRYuXCiPx/qP6NFHH9VLL73U4fd99913tWTJEq1YsUKTJk2SJJ1wwgmh88FtbwMHDlRxcXG7z7F161YtW7ZMb7zxhiZMmCBJeuaZZ+RyubR06VJ95StfkSQdPnxYjz76qE488URJ0rx583T33XdLkpqbm9XU1KQvf/nLofOjR4/u/BsJAACQJIGWgLwNXvn3+eXo65B7iDvqSo5ntEflI8s7fR06jySoI15v5ApQa6YpNTZacWVlCZnCiy++qD59+ujw4cNqaWnRzJkzdeedd2ru3Lk69dRTw+4D+stf/qJt27apb9++Yc9x4MABbd++XU1NTfL7/Ro3blzoXI8ePXT22WdHbIkL2rhxo+x2u84777wuv4a6ujr16NEj7Pv2799fI0eOVF1dXWisd+/eoQRHkhwOh/bs2SPJSrZmz56tyZMn60tf+pImTZqk6dOny+Hg5kAAAJB+jDpD85fPDyt77Sx0qnpKddQVHXueXWXDyhI8Q7AdriP+GEsQxhrXBRMnTtTGjRu1detWffrpp3r66ad17LHHSlLoa9DHH3+ssWPHauPGjWGPd999VzNnzuzS9+/Vq1e3X0OsevbsGXZss9nCkrOamhqtWbNGEyZM0PPPP6+TTz5Za9euTdr8AAAAYmHUGapYUhHR98fX7FPFkgoZdcm9nQLtIwnqSKyrDAlcjTj22GM1YsQIDRkyRD16HH3R7qyzztLWrVs1cOBAjRgxIuxRVFSkoqIiORwOrVu3LnTNkSNHtGHDhg6f89RTT1VLS4teffXVds8HV6ICgY67F48ePVpHjhwJ+74ffPCBtmzZos997nNHfU1tnXnmmVq4cKHefPNNjRkzRs8++2ynrgcAAEikQEtA85fPb7fCW3CscnmlAi0d/+6E5CAJ6ojbLTmd1r0/7bHZJJfLiksDV1xxhY4//niVl5fL6/Wqvr5eq1ev1g033KD3P9vWN3/+fN1zzz1aunSpNm/erOuuuy6ix09rw4YN06xZs/S1r31NS5cuDT3nkiVLJElDhw6VzWbTiy++qH/+85/6+OOPI57jpJNOUnl5uebMmaPXX39df/nLX/TVr35VpaWlKi8vj+m11dfXa+HChVqzZo3ee+89/fGPf9TWrVu5LwgAAKQVb4M3YgWoNVOmGpsb5W3wJnFWaA9JUEfsdilY7axtIhQ8rqpKSr+gWPTu3VuvvfaahgwZIo/Ho9GjR+uaa67RgQMHVFhYKEm66aabdOWVV2rWrFkaP368+vbtq2nTph31eX/605+qoqJC1113nUaNGqU5c+Zo//79kqTS0lLddddduvXWWzVo0CDNmzev3eeoqanR2LFj9eUvf1njx4+XaZr6/e9/H7EF7mivbfPmzbrssst08skn69prr9XcuXP1jW98oxPvEAAAQGL598V2m0SscUgcm9nRXfEZorm5WUVFRWpqagr9sh904MAB1dfXa/jw4TrmmGO69g0Mw6oS17pIgstlJUAeShVmm7h8ZgAAQE5avWO1Jj49MWrcK7NeofhBghwtN2iN6nDReDxWGWyv1yqC4HBYW+DSZAUIAAAA6cE9xC1noVO+Zl+79wXZZJOz0Cn3kPS4nSKXkQTFwm5PWBlsAAAAZAd7nl3VU6pVsaRCNtnCEiGbrNspqqZU0fcnDXBPEAAAABAnntEe1U6vVWlhadi4s9Cp2um1UfsEITlYCQIAAAA6EGgJyNvglX+fX46+DrmHuKOu5HhGe1Q+srzT1yF5ciIJyvDaD0giPisAACDIqDM0f/n8sLLXzkKnqqdUR13RsefZKX6QxrJ6O1ywBPMnn3yS4pkgUwQ/K7GW7wYAANnJqDNUsaQiou+Pr9mniiUVMuqMFM0M8ZDVK0F2u13FxcXas2ePJKvfjK2j5qfIaaZp6pNPPtGePXtUXFwsO9X/AADIWYGWgOYvn99uhTdTpmyyqXJ5pcpHlrPFLUNldRIkSSUlJZIUSoSAoykuLg59ZgAAQG7yNngjVoBaM2WqsblR3gYvW94yVNYnQTabTQ6HQwMHDtThw4dTPR2ksZ49e7ICBAAA5N/nj2sc0k/WJ0FBdrudX3ABAAAQlaOvI65xSD9ZXRgBAAAA6Cz3ELechc5Qg9O2bLLJVeiSe4g7yTNDvJAEAQAAAK3Y8+yqnlItSRGJUPC4akoVRREyGEkQAAAA0IZntEe102tVWlgaNu4sdKp2em3UPkFIbzYzw7tDNjc3q6ioSE1NTSosLEz1dAAAAJBmAi0BeRu88u/zy9HXIfcQd8yrON25FskXa26QM4URAAAAkHuMOkPzl88PK3ntLHSqekp1TKs59jw7ZbCzENvhAAAAkJWMOkMVSyoiev74mn2qWFIho85I0cyQaiRBAAAAyDqBloDmL58vU5F3fgTHKpdXKtASSPbUkAZIggAAAJB1vA3eiBWg1kyZamxulLfBm8RZIV2QBAEAACDr+Pf54xqH7EISBAAAgKzj6OuIaxyyC0kQAAAAso57iFvOQmdEs9Mgm2xyFbrkHuJO8syQDkiCAAAAkHXseXZVT6mWpIhEKHhcNaWKnj85iiQIAAAAWckz2qPa6bUqLSwNG3cWOlU7vTamPkHITjbTNCPrBmaQWLvCAgAAILMFWgLyNnjl3+eXo69D7iHumFZyunodMk+suUGPJM4JAAAA6BKjztD85fPDyl47C52qnlIddUXHnmdX2bCyBM8QmYTtcAAAAEhrRp2hiiUVEX1/fM0+VSypkFFnpGhmyFQkQQAAAEhbgZaA5i+fL1ORd3AExyqXVyrQEkj21JDBSIIAAACQtrwN3ogVoNZMmWpsbpS3wZvEWSHTkQQBAAAgbfn3+eMaB0gkQQAAAEhjjr6OuMYBEkkQAAAA0ph7iFvOQmdEw9Mgm2xyFbrkHuJO8syQyRKaBA0bNkw2my3iMXfuXElSWVlZxLlvfvObiZwSAAAAMog9z67qKdWSFJEIBY+rplTR9wedktAkaP369fL7/aHHihUrJElf+cpXQjFz5swJi7n33nsTOSUAAABkGM9oj2qn16q0sDRs3FnoVO302qh9goC2EtosdcCAAWHH99xzj0488USdd955obHevXurpKQkkdMAAABAGgm0BORt8Mq/zy9HX4fcQ9xRV3I8oz0qH1ne6euA9iQ0CWrt0KFD+sUvfqEFCxbIZvv3UuYzzzyjX/ziFyopKdEll1yi7373u+rdu3eHz3Pw4EEdPHgwdNzc3JzQeQMAACB+jDpD85fPDyt77Sx0qnpKddQVHXueXWXDyhI8Q+SCpCVBS5cu1d69ezV79uzQ2MyZMzV06FANHjxYmzZt0i233KItW7bIMDru+rto0SLdddddSZgxAAAA4smoM1SxpCKi8amv2aeKJRVsbUPS2EzTjGy/mwCTJ09Wfn6+fvvb33YY8/LLL+v888/Xtm3bdOKJJ7Yb095KkMvlUlNTkwoLC+M+bwAAAHRfoCWgYdXDOmx8apNNzkKn6ufXs8UNXdbc3KyioqKouUFSSmS/9957Wrlypb7+9a8fNW7cuHGSpG3btnUYU1BQoMLCwrAHAAAA0pu3wdthAiRJpkw1NjfK2+BN4qyQq5KSBNXU1GjgwIG6+OKLjxq3ceNGSZLDQbMrAACAbOLf549rHNAdCb8nqKWlRTU1NZo1a5Z69Pj3t9u+fbueffZZXXTRRerfv782bdqkG2+8Ueeee65OO+20RE8LAAAASeToG9sfuWONA7oj4UnQypUr1dDQoK997Wth4/n5+Vq5cqWqqqq0f/9+uVwuXXbZZbrtttsSPSUAAAAkmXuIW85Cp3zNvojCCNK/7wlyD3GnYHbINUkrjJAosd78BAAAgNQKVoeTFJYI2WS1T6E6HLorrQojAAAAAJ7RHtVOr1VpYWnYuLPQSQKEpGIlCAAAAJ0WaAnI2+CVf59fjr4OuYe4Yy5t3Z1rgaOJNTdIWrNUAAAAZAejztD85fPDSl47C52qnlId02qOPc+usmFlCZwhcHRshwMAAEDMgvf1tO3542v2qWJJhYw6I0UzA2JHEgQAAICYBFoCmr98frvV3YJjlcsrFWgJJHtqQKeQBAEAACAm3gZvxApQa6ZMNTY3ytvgTeKsgM4jCQIAAEBM/Pv8cY0DUoUkCAAAADFx9HXENQ5IFZIgAAAAxMQ9xC1noTPU3LQtm2xyFbrkHuJO8syAziEJAgAAQEzseXZVT6mWpIhEKHhcNaWKnj9IeyRBAAAAiJlntEe102tVWlgaNu4sdKp2em1MfYKAVLOZphlZ4zCDxNoVFgAAAJECAcnrlfx+yeGQ3G7JHsNCTqAlIG+DV/59fjn6OuQe4mYFCCkXa27QI4lzAgAAQBoxDGn+fOn9VlWvnU6pulryRFnQsefZVTasLKHzAxKF7XAAAAA5yDCkiorwBEiSfD5r3DBSMy8gGUiCAAAAckwgYK0AtXdTRHCsstKKA7IRSRAAAECO8XojV4BaM02psdGKA7IRSRAAAECO8fvjGwdkGpIgAACAHONwxDcOyDQkQQAAADnG7baqwNls7Z+32SSXy4oDshFJEAAAQI6x260y2FJkIhQ8rqqKrV8QkIlIggAAAHKQxyPV1kqlpeHjTqc1Hq1PEJDJaJYKAACQ4QIBq5Kb32/dx+N2x7aK4/FI5eVduxbIZCRBAAAAGcwwrJ4/rUteO53WdrdYVnPsdqmsLGHTA9IS2+EAAAAylGFIFRWRPX98PmvcMFIzLyDdkQQBAABkoEDAWgEyzchzwbHKSisOQDiSIAAAgAzk9UauALVmmlJjoxUHIBxJEAAAQAby++MbB+QSkiAAAIAM5HDENw7IJSRBAAAAGcjttqrAtW12GmSzSS6XFQcgHEkQAABABrLbrTLYUmQiFDyuqqLnD9AekiAAAIAM5fFItbVSaWn4uNNpjcfSJwjIRTRLBQAASBOBgFXNze+37uVxu6Ov5Hg8Unl5568DchlJEAAAQBowDKvvT+uy106nteUt2oqO3S6VlSV0ekBWYTscAABAihmGVFER2ffH57PGDSM18wKyFUkQAABACgUC1gqQaUaeC45VVlpxAOKDJAgAACCFvN7IFaDWTFNqbLTiAMQHSRAAAEAK+f3xjQMQHUkQAABACjkc8Y0DEB1JEAAAQAq53VYVuLYNT4NsNsnlsuIAxAdJEAAAQArZ7VYZbCkyEQoeV1XR9weIJ5IgAACAFPN4pNpaqbQ0fNzptMaj9QkC0Dk0SwUAAIizQMCq5ub3W/fyuN3RV3I8Hqm8vPPXAeg8kiAAAIA4Mgyr70/rstdOp7XlLdqKjt0ulZUldHoAxHY4AACAuDEMqaIisu+Pz2eNG0Zq5gUgHEkQAABAHAQC1gqQaUaeC45VVlpxAFKLJAgAACAOvN7IFaDWTFNqbLTiAKQWSRAAAEAc+P3xjQOQOCRBAAAAceBwxDcOQOKQBAEAAMSB221VgWvb8DTIZpNcLisOQGqRBAEAAMSB3W6VwZYiE6HgcVUVfX+AdEASBAAAECcej1RbK5WWho87ndZ4tD5BAJKDZqkAAADtCASsSm5+v3Ufj9sd2yqOxyOVl3ftWgDJkdCVoDvvvFM2my3sMWrUqND5AwcOaO7cuerfv7/69Omjyy67TLt3707klAAAAKIyDGnYMGniRGnmTOvrsGGxNzu126WyMmnGDOsrCRCQXhK+He6UU06R3+8PPV5//fXQuRtvvFG//e1v9cILL+jVV1/Vzp075WGdGAAApJBhSBUVkT1/fD5rPNZECED6Svh2uB49eqikpCRivKmpSU8++aSeffZZ/ed//qckqaamRqNHj9batWv1H//xH4meGgAAQJhAQJo/32ps2pZpWgUOKiut7W6s7gCZK+ErQVu3btXgwYN1wgkn6IorrlBDQ4MkacOGDTp8+LAmTZoUih01apSGDBmiNWvWdPh8Bw8eVHNzc9gDAAAgHrzeyBWg1kxTamy04gBkroQmQePGjdPixYu1fPly/fSnP1V9fb3cbrf27dunXbt2KT8/X8XFxWHXDBo0SLt27erwORctWqSioqLQw+VyJfIlAACAHOL3xzcOQHpK6Ha4Cy+8MPTv0047TePGjdPQoUO1ZMkS9erVq0vPuXDhQi1YsCB03NzcTCIEAADiwuGIbxyA9JTUPkHFxcU6+eSTtW3bNpWUlOjQoUPau3dvWMzu3bvbvYcoqKCgQIWFhWEPAACAeHC7rZ4+bZudBtlskstlxQHIXElNgj7++GNt375dDodDY8eOVc+ePbVq1arQ+S1btqihoUHjx49P5rQAAAAkWcUOqqutf7dNhILHVVUURQAyXUKToJtvvlmvvvqqduzYoTfffFPTpk2T3W7XjBkzVFRUpGuuuUYLFizQK6+8og0bNujqq6/W+PHjqQwHAABSxuORamul0tLwcafTGqebB5D5EnpP0Pvvv68ZM2bogw8+0IABA3TOOedo7dq1GjBggCTpRz/6kfLy8nTZZZfp4MGDmjx5sn7yk58kckoAACDHBAJWNTe/37qXx+2OvpLj8VhlsDt7HYDMYDPN9irhZ47m5mYVFRWpqamJ+4MAAEAYw7D6/rQue+10WlveWNEBsk+suUFS7wkCAABIFsOQKioi+/74fNa4YaRmXgBSjyQIAABknUDAWgFqb79LcKyy0ooDkHtIggAAQNbxeiNXgFozTamx0YoDkHtIggAAQNbx++MbByC7kAQBAICs43DENw5AdiEJAgAAWcfttqrAtW14GmSzSS6XFQcg95AEAQCArGO3W2WwpchEKHhcVUXfHyBXkQQBAICs5PFItbVSaWn4uNNpjdMnCMhdPVI9AQAAgFgEAlY1N7/fupfH7Y6+kuPxSOXlnb8OQHYjCQIAAGnPMKy+P63LXjud1pa3aCs6drtUVpbQ6QHIMGyHAwAAac0wpIqKyL4/Pp81bhipmReAzEUSBAAA0lYgYK0AmWbkueBYZaUVBwCxIgkCAABpy+uNXAFqzTSlxkYrDgBiRRIEAADSlt8f3zgAkEiCAABAGnM44hsHABJJEAAASGNut1UFrm3D0yCbTXK5rDgAiBVJEAAASFt2u1UGW4pMhILHVVX0/QHQOSRBAAAgrXk8Um2tVFoaPu50WuPR+gQBQFs0SwUAAEkTCFiV3Px+6z4etzu2VRyPRyov79q1ANAWSRAAAEgKw7B6/rQuee10WtvdYlnNsdulsrKETQ9ADmE7HAAASDjDkCoqInv++HzWuGGkZl4AchNJEAAASKhAwFoBMs3Ic8GxykorDgCSgSQIAAAklNcbuQLUmmlKjY1WHAAkA0kQAABIKL8/vnEA0F0kQQAAIKEcjvjGAUB3kQQBAICEcrutKnBtm50G2WySy2XFAUAykAQBAICEstutMthSZCIUPK6qoucPgOQhCQIAAAnn8Ui1tVJpafi402mNx9InCADihWapAACg0wIBq5qb32/dy+N2R1/J8Xik8vLOXwcA8UYSBAAAOsUwrL4/rcteO53WlrdoKzp2u1RWltDpAUBUbIcDAAAxMwypoiKy74/PZ40bRmrmBQCdQRIEAABiEghYK0CmGXkuOFZZacUBQDojCQIAADHxeiNXgFozTamx0YoDgHRGEgQAAGLi98c3DgBShSQIAADExOGIbxwApApJEAAAiInbbVWBa9vwNMhmk1wuKw4A0hlJEAAAiIndbpXBliIToeBxVRV9fwCkP5IgAAAQM49Hqq2VSkvDx51OazxanyAASAc0SwUAIEcFAlYlN7/fuo/H7Y5tFcfjkcrLu3YtAKQDkiAAAHKQYVg9f1qXvHY6re1usazm2O1SWVnCpgcACcV2OAAAcoxhSBUVkT1/fD5r3DBSMy8ASBaSIAAAckggYK0AmWbkueBYZaUVBwDZiiQIAIAc4vVGrgC1ZppSY6MVBwDZiiQIAIAc4vfHNw4AMhFJEAAAOcThiG8cAGQikiAAAHKI221VgWvb7DTIZpNcLisOALIVSRAAADnEbrfKYEuRiVDwuKqKnj8AshtJEAAAOcbjkWprpdLS8HGn0xqPpU8QAGQymqUCAJDhAgGrmpvfb93L43ZHX8nxeKTy8s5fBwDZgCQIAIAMZhhW35/WZa+dTmvLW7QVHbtdKitL6PQAIC2xHQ4AgAxlGFJFRWTfH5/PGjeM1MwLANJdQpOgRYsW6fOf/7z69u2rgQMHaurUqdqyZUtYTFlZmWw2W9jjm9/8ZiKnBQBAxgsErBUg04w8FxyrrLTiAADhEpoEvfrqq5o7d67Wrl2rFStW6PDhw7rgggu0f//+sLg5c+bI7/eHHvfee28ipwUAQMbzeiNXgFozTamx0YoDAIRL6D1By5cvDztevHixBg4cqA0bNujcc88Njffu3VslJSWJnAoAAFnF749vHADkkqTeE9TU1CRJ6tevX9j4M888o+OPP15jxozRwoUL9cknn3T4HAcPHlRzc3PYAwCAXONwxDcOAHJJ0qrDtbS0qLKyUl/84hc1ZsyY0PjMmTM1dOhQDR48WJs2bdItt9yiLVu2yOjgbs5FixbprrvuSta0AQBIS263VQXO52v/viCbzTrvdid/bgCQ7mym2d6Pzvj71re+pT/84Q96/fXX5XQ6O4x7+eWXdf7552vbtm068cQTI84fPHhQBw8eDB03NzfL5XKpqalJhYWFCZk7AADpKFgdTgpPhGw26yuNTwEkXFcalSVQc3OzioqKouYGSdkON2/ePL344ot65ZVXjpoASdK4ceMkSdu2bWv3fEFBgQoLC8MeAADkIo/HSnRKS8PHnU4SIABJYBjSsGHSxInSzJnW12HDMqI+f0K3w5mmqeuvv16//vWvtXr1ag0fPjzqNRs3bpQkOdjEDADIMV35g6rHI5WXp9UfYgFkmq788AkuRbfdVBZsVJbmf4lJ6Ha46667Ts8++6x+85vfaOTIkaHxoqIi9erVS9u3b9ezzz6riy66SP3799emTZt04403yul06tVXX43pe8S65AUAQDozDKvvT+uy106nVF2d1r9HAMh0XfnhEwhYKz4d1ekP3pRYX5/0v8jEmhskNAmyBTclt1FTU6PZs2ersbFRX/3qV/XOO+9o//79crlcmjZtmm677baYExqSIABApuvoD6rc2wOgUzq7otPVHz6rV1tb36J55RWprCzW2cdFWiRByUASBADIZGn8B1UAmaSzKzrd+eHz3HPWPUDRPPusNGNGzC8hHtKqMAIAAGif19vx7yCS9QfaxkYrDkAOCASslZbnnrO+BgLRrwmu6LT9YRK8P6e9QgXd+eGTBY3KSIIAAEghvz++cQAyWFeqrQUC1gpQe5u7gmOVlZHJVHd++AQblXVw64tsNsnlSutGZSRBAACkUBb8QRVAezq7otOV1Ryp6ys63fnhY7db2+ykyEQoeFxVldZ7eEmCAABIoSz4gyqQ3bq6Pa0zKzpdXc2Rur6i090fPhneqIwkCACAFMqCP6gC2asr29My5f6cePzw8XikHTusKnDPPmt9ra9P+wRIIgkCACDlMvwPqkD6S1axgUy7PyceP3zsdqsM9owZ1tcM+YsNJbIBAIijrjRej8e1ADqQzGagXe2f092+O8GETQpPwGJtNpZFP3xizQ16JHFOAABkta78rtVa8A+qADoQr2agwdWcjpKDzmxPa/0fbXfvz/H52l9FCiZd0e7Pae8HUFVV9B9AOfjDh+1wAADEQVcLOwGIUSYUG+D+nIxBEgQAQDd153ctICclo3x0KooNcH9OxiAJAgCgm7rzuxaQc5K1opOKYgPdXdFhNSdpSIIAAOim7vyuBWSsZFVcy7RmoN1d0WE1JylIggAA6Kbu/K4FZKSu9M9J9opOKpuBsqKT9qgOBwBAN3W3sBOQUZJdca27xQYqKqz/CNsrHR1LsYHy8q6Vj87BimuZhJUgAAC6KR6FnYCMkIqKaxQbQAKQBAEAEAfx+F0LSHupqLhGsQEkANvhAABoR1caqHdn5wyQEeJRca0r+0ZpBoo4IwkCAKANw2j/d63qan7XQo6LR8W1rt6jw18ZEEc202wvFc8czc3NKioqUlNTkwoLC1M9HQBAhuvonu/g72hsbUNOCwSsKnDRVnPq6ztOTtr7K4PLFduKDhBFrLkBSRAAAJ8J/n7X0S0Psfx+B2S94F8KpPZXc2L5S0FX9psCMYg1N6AwAgAAn+nOPd9AzqDiGrIA9wQBAPCZ7tzzDeQU7s9BhiMJAgDgM9255xvIOVQBQQZjOxwAAJ/pTk9GAEDmIAkCAOAz3e3JCADIDCRBAAC0Eo97vgEA6Y17ggAAWa0rlXi55xsAshtJEAAga7XXk9HptLa8RVvR4Z5vAMhebIcDAGSlYD/Htn1/fD5r3DBSMy8AQOqRBAEAsk4gYK0AtW5mHxQcq6y04gAAuYckCACQdbzeyBWg1kxTamy04gAAuYckCACQdfz++MYBALILSRAAIOs4HPGNAwBkF5IgAEDWcbutKnBtG54G2WySy2XFAQByD0kQACDr2O1WGWwpMhEKHldV0fcHAHIVSRAAICt5PFJtrVRaGj7udFrj0foEAQCyF81SAQBpLxCwKrn5/dZ9PG53bKs4Ho9UXt61awEA2YskCACQ1gzD6vnTuuS102ltd4tlNcdul8rKEjY9AEAGYjscACBtGYZUURHZ88fns8YNIzXzAgBkNpIgAEBaCgSsFSDTjDwXHKustOIAAOgMkiAAQFryeiNXgFozTamx0YoDAKAzSIIAAGnJ749vHAAAQSRBAIC05HDENw4AgCCSIABAWnK7rSpwbZudBtlskstlxQEA0BkkQQCAtGS3W2WwpchEKHhcVUXPHwBA55EEAQDSlscj1dZKpaXh406nNR5LnyAAANqiWSoAIGkCAauam99v3cvjdkdfyfF4pPLyzl8HAEBHSIIAAElhGFbfn9Zlr51Oa8tbtBUdu10qK0vo9AAAOYTtcACAhDMMqaIisu+Pz2eNG0Zq5gUAyE0kQQCAhAoErBUg04w8FxyrrLTiAABIBpIgAEBCeb2RK0CtmabU2GjFAQCQDCRBAICE8vvjGwcAQHelRRL0yCOPaNiwYTrmmGM0btw4vfXWW6meEgAgThyO+MYBANBdKU+Cnn/+eS1YsEB33HGH/vznP+v000/X5MmTtWfPnlRPDQAQB263VQWubcPTIJtNcrmsOAAAkiHlSdCDDz6oOXPm6Oqrr9bnPvc5Pfroo+rdu7eeeuqpVE8NABAHdrtVBluKTISCx1VV9P0BACRPSpOgQ4cOacOGDZo0aVJoLC8vT5MmTdKaNWvavebgwYNqbm4OewAA0pvHI9XWSqWl4eNOpzUerU8QAADxlNJmqf/6178UCAQ0aNCgsPFBgwZp8+bN7V6zaNEi3XXXXcmYHgCgHYGAVcnN77fu43G7Y1vF8Xik8vKuXQsAQDylNAnqioULF2rBggWh4+bmZrlcrhTOCAByh2FYPX9al7x2Oq3tbrGs5tjtUllZwqYHAEBMUpoEHX/88bLb7dq9e3fY+O7du1VSUtLuNQUFBSooKEjG9AAArRiGVFER2fTU57PG2dYGAMgUKb0nKD8/X2PHjtWqVatCYy0tLVq1apXGjx+fwpkBAFoLBKwVoLYJkPTvscpKKw4AgHSX8upwCxYs0M9+9jM9/fTTqqur07e+9S3t379fV199daqnBgD4jNcbvgWuLdOUGhutOAAA0l3K7wm6/PLL9c9//lO33367du3apTPOOEPLly+PKJYAAEgdvz++cQAApFLKkyBJmjdvnubNm5fqaQAAOuBwxDcOAIBUSvl2OABA+nO7rSpwbZudBtlskstlxQEAkO5IggAAUdntVhlsKTIRCh5XVdHzBwCQGUiCAAAx8XisMtilpeHjTiflsQEAmSUt7gkCACRfIGBVc/P7rXt53O7oKzkej1Re3vnrAABIJyRBAJCDDMPq+9O67LXTaW15i7aiY7dLZWUJnR4AAAnFdjgAyDGGIVVURPb98fmsccNIzbwAAEgWkiAAyCGBgLUCZJqR54JjlZVWHAAA2YokCAByiNcbuQLUmmlKjY1WHAAA2YokCAByiN8f3zgAADIRSRAA5BCHI75xAABkIpIgAMghbrdVBa5tw9Mgm01yuaw4AACyFUkQAOQQu90qgy1FJkLB46oq+v4AALIbSRAA5BiPR6qtlUpLw8edTms8Wp8gAAAyHc1SASDDBQJWNTe/37qXx+2OvpLj8Ujl5Z2/DgCAbEASBAAZzDCsvj+ty147ndaWt2grOna7VFaW0OkBAJCW2A4HABnKMKSKisi+Pz6fNW4YqZkXAADpjiQIADJQIGCtAJlm5LngWGWlFQcAAMKRBAFABvJ6I1eAWjNNqbHRigMAAOFIggAgA/n98Y0DACCXkAQBQAZyOOIbBwBALiEJAoAM5HZbVeDaNjwNstkkl8uKAwAA4UiCACAD2e1WGWwpMhEKHldV0fcHAID2kAQBQIbyeKTaWqm0NHzc6bTGo/UJAgAgV9EsFQDSQCBgVXLz+637eNzu2FZxPB6pvLxr1wIAkKtIggAgxQzD6vnTuuS102ltd4tlNcdul8rKEjY9AACyDtvhACCFDEOqqIjs+ePzWeOGkZp5AQCQzUiCACBFAgFrBcg0I88FxyorrTgAABA/JEEAkCJeb+QKUGumKTU2WnEAACB+SIIAIEX8/vjGAQCA2JAEAUCKOBzxjQMAALEhCQKAFHG7rSpwbZudBtlskstlxQEAgPghCQKAFLHbrTLYUmQiFDyuqqLnDwAA8UYSBAAp5PFItbVSaWn4uNNpjcfSJwgAAHQOzVIBII4CAauam99v3cvjdkdfyfF4pPLyzl8HAAC6hiQIAOLEMKy+P63LXjud1pa3aCs6drtUVpbQ6QEAgM+wHQ4A4sAwpIqKyL4/Pp81bhipmRcAAIhEEgQA3RQIWCtAphl5LjhWWWnFAQCA1CMJAoBu8nojV4BaM02psdGKAwAAqUcSBADd5PfHNw4AACQWSRAAdJPDEd84AACQWCRBANBNbrdVBa5tw9Mgm01yuaw4AACQeiRBANBNdrtVBluKTISCx1VV9P0BACBdkAQBQBx4PFJtrVRaGj7udFrj0foEAQCA5KFZKgC0IxCwqrn5/da9PG539JUcj0cqL+/8dQAAILlIggCgDcOw+v60LnvtdFpb3qKt6NjtUllZQqcHAAC6ie1wANCKYUgVFZF9f3w+a9wwUjMvAAAQPyRBAPCZQMBaATLNyHPBscpKKw4AAGQukiAA+IzXG7kC1JppSo2NVhwAAMhcJEEA8Bm/P75xAAAgPZEEAcBnHI74xgEAgPREEgQAn3G7rSpwbRueBtlskstlxQEAgMyVkCRox44duuaaazR8+HD16tVLJ554ou644w4dOnQoLMZms0U81q5dm4gpAUBUdrtVBluKTISCx1VV9P0BACDTJaRP0ObNm9XS0qLHHntMI0aM0DvvvKM5c+Zo//79uv/++8NiV65cqVNOOSV03L9//0RMCQBi4vFItbXt9wmqqoreJwgAAKQ/m2m2Vww2/u677z799Kc/1T/+8Q9J1krQ8OHD9fbbb+uMM87o8vM2NzerqKhITU1NKiwsjNNsAWSDQMCq5Ob3W/fxuN2xr+J051oAAJAaseYGCVkJak9TU5P69esXMX7ppZfqwIEDOvnkk/Wd73xHl1566VGf5+DBgzp48GDouLm5Oe5zBZD5DKP91Zzq6thWc+x2qawsYdMDAAAplJTCCNu2bdPDDz+sb3zjG6GxPn366IEHHtALL7yg3/3udzrnnHM0depULVu27KjPtWjRIhUVFYUeLpcr0dMHkGEMQ6qoiOz54/NZ44aRmnkBAID00KntcLfeeqt++MMfHjWmrq5Oo0aNCh37fD6dd955Kisr0xNPPHHUa6+66irV19fLe5ROhO2tBLlcLrbDAZBkbWMbNqzjpqc2m7UiVF/P9jYAALJNQrbD3XTTTZo9e/ZRY0444YTQv3fu3KmJEydqwoQJevzxx6M+/7hx47RixYqjxhQUFKigoCCm+QLIPV5vxwmQJJmm1NhoxbHdDQCA3NSpJGjAgAEaMGBATLE+n08TJ07U2LFjVVNTo7y86DvvNm7cKAddCAF0g98f3zgAAJB9ElIYwefzqaysTEOHDtX999+vf/7zn6FzJSUlkqSnn35a+fn5OvPMMyVJhmHoqaeeirplDgCOJta/o/D3FgAAcldCkqAVK1Zo27Zt2rZtm5xOZ9i51rcgfe9739N7772nHj16aNSoUXr++edVUVGRiCkByBFut3XPj89nbX1rK3hPkNud/LkBAID0kLQ+QYlCnyAAbQWrw0nhiZDNZn2traXpKQAA2SjW3CApJbIBIJk8HivRKS0NH3c6SYAAAEASm6UCQFcFAlY1N7/fupfH7Y5e3trjkcrLO38dAADIfiRBANKaYUjz54eXvXY6perq6Cs6djtlsAEAQCS2wwFIW8F7e9r2/fH5rHHDSM28AABAZiMJApCWAgFrBai90i3BscpKKw4AAKAzSIIApCWvN3IFqDXTlBobrTgAAIDOIAkCkJb8/vjGAQAABJEEAUhLDkd84wAAAIJIggCkJbfbqgIXbHDals0muVxWHAAAQGeQBAFIS3a7VQZbikyEgsdVVfT9AQAAnUcSBCBteTxSba1UWho+7nRa49H6BAEAALSHZqkAkiIQsCq5+f3WfTxud2yrOB6PVF7etWsBAADaQxIEIOEMw+r507rktdNpbXeLZTXHbpfKyhI2PQAAkGPYDgcgoQxDqqiI7Pnj81njhpGaeQEAgNxFEgQgYQIBawXINCPPBccqK604AACAZCEJApAwXm/kClBrpik1NlpxAAAAyUISBCBh/P74xgEAAMQDSRCAhHE44hsHAAAQDyRBABLG7baqwLVtdhpks0kulxUHAACQLCRBABLGbrfKYEuRiVDwuKqKnj8AACC5SIIAJJTHI9XWSqWl4eNOpzUeS58gAACAeKJZKoBOCQSsam5+v3Uvj9sdfSXH45HKyzt/HQAAQCKQBAGImWFYfX9al712Oq0tb9FWdOx2qawsodMDAACICdvhAMTEMKSKisi+Pz6fNW4YqZkXAABAZ5EEAYgqELBWgEwz8lxwrLLSigMAAEh3JEEAovJ6I1eAWjNNqbHRigMAAEh3JEEAovL74xsHAACQSiRBAKJyOOIbBwAAkEokQQCicrutKnBtG54G2WySy2XFAQAApDuSIABR2e1WGWwpMhEKHldV0fcHAABkBpIgADHxeKTaWqm0NHzc6bTGo/UJAgAASBc0SwVyVCBgVXPz+617edzu6Cs5Ho9UXt756wAAANIJSRCQgwzD6vvTuuy102lteYu2omO3S2VlCZ0eAABAQrEdDsgxhiFVVET2/fH5rHHDSM28AAAAkoUkCMghgYC1AmSakeeCY5WVVhwAAEC2IgkCcojXG7kC1JppSo2NVhwAAEC2IgkCcojfH984AACATEQSBOQQhyO+cQAAAJmIJAjIIW63VQWubcPTIJtNcrmsOAAAgGxFEgTkELvdKoMtRSZCweOqKvr+AACA7EYSBOQYj0eqrZVKS8PHnU5rPFqfIAAAgExHs1QggwUCViU3v9+6j8ftjm0Vx+ORysu7di0AAECmIwkCMpRhWD1/Wpe8djqt7W6xrObY7VJZWcKmBwAAkLbYDgdkIMOQKioie/74fNa4YaRmXgAAAJmAJAjIMIGAtQJkmpHngmOVlVYcAAAAIpEEARnG641cAWrNNKXGRisOAAAAkUiCgAzj98c3DgAAINeQBAEZxuGIbxwAAECuIQkCMozbbVWBa9vsNMhmk1wuKw4AAACRSIKADGO3W2WwpchEKHhcVUXPHwAAgI6QBAEZyOORamul0tLwcafTGo+lTxAAAECuolkqkAYCAauam99v3cvjdkdfyfF4pPLyzl8HAACQ6xK2EjRs2DDZbLawxz333BMWs2nTJrndbh1zzDFyuVy69957EzUdIG0ZhjRsmDRxojRzpvV12LDYGp7a7VJZmTRjhvWVBAgAACC6hK4E3X333ZozZ07ouG/fvqF/Nzc364ILLtCkSZP06KOP6q9//au+9rWvqbi4WNdee20ipwWkDcOQKioiG5/6fNY4W9sAAADiL6FJUN++fVVSUtLuuWeeeUaHDh3SU089pfz8fJ1yyinauHGjHnzwQZIg5IRAQJo/PzIBkqwxm02qrLS2vLHCAwAAED8JLYxwzz33qH///jrzzDN133336ciRI6Fza9as0bnnnqv8/PzQ2OTJk7VlyxZ99NFHHT7nwYMH1dzcHPYAMpHXK73/fsfnTVNqbLTiAAAAED8JWwm64YYbdNZZZ6lfv3568803tXDhQvn9fj344IOSpF27dmn48OFh1wwaNCh07rjjjmv3eRctWqS77rorUdMGksbvj28cAAAAYtOplaBbb701othB28fmzZslSQsWLFBZWZlOO+00ffOb39QDDzyghx9+WAcPHuzWhBcuXKimpqbQo7GxsVvPB6SKwxHfOAAAAMSmUytBN910k2bPnn3UmBNOOKHd8XHjxunIkSPasWOHRo4cqZKSEu3evTssJnjc0X1EklRQUKCCgoLOTBtIS2631dfH52v/viCbzTrvdid/bgAAANmsU0nQgAEDNGDAgC59o40bNyovL08DBw6UJI0fP17/8z//o8OHD6tnz56SpBUrVmjkyJEdboUDsondLlVXW1XgbLbwRMhms75WVVEUAQAAIN4SUhhhzZo1qqqq0l/+8hf94x//0DPPPKMbb7xRX/3qV0MJzsyZM5Wfn69rrrlGf/vb3/T888+rurpaCxYsSMSUgLTk8VhlsEtLw8edTspjAwAAJIrNNNvbiNM9f/7zn3Xddddp8+bNOnjwoIYPH64rr7xSCxYsCNvKtmnTJs2dO1fr16/X8ccfr+uvv1633HJLp75Xc3OzioqK1NTUpMLCwni/FCBmgYBVyc3vt+7jcbtjX8XpzrUAAACwxJobJCQJSiaSIKQDw7B6/rQuee10WtvdWM0BAABIjlhzg4T2CQJygWFY9/W07fnj81njhpGaeQEAAKB9JEFANwQC1gpQe+upwbHKSisOAAAA6YEkCOgGrzdyBag105QaG604AAAApAeSIKAb/P74xgEAACDxSIKAbnA44hsHAACAxCMJArrB7baqwAWbm7Zls0kulxUHAACA9EASBHSD3W6VwZYiE6HgcVUVPX8AAADSCUkQ0E0ej1RbK5WWho87ndY4fYIAAADSS49UTwBIN4GAVc3N77fu5XG7o6/keDxSeXnnrwMAAEDykQQBrRiG1fenddlrp9Pa8hZtRcdul8rKEjo9AAAAxAHb4YDPGIZUURHZ98fns8YNIzXzAgAAQHyRBAGytsDNn281N20rOFZZacUBAAAgs5EEAbLu5Wm7AtSaaUqNjVYcAAAAMhtJECCrmEE84wAAAJC+SIIAWdXc4hkHAACA9EUSBMgqZ+10RjY8DbLZJJfLigMAAEBmIwkCZJW3rq62/t02EQoeV1XR9wcAACAbkAQBn/F4pNpaqbQ0fNzptMaj9QkCAABAZqBZKrJWIGBVc/P7rXt53O7oKzkej1Re3vnrAAAAkDlIgpCVDMPq+9O67LXTaW15i7aiY7dLZWUJnR4AAABSiO1wyDqGIVVURPb98fmsccNIzbwAAACQHkiCkFUCAWsFyDQjzwXHKiutOAAAAOQmkiBkFa83cgWoNdOUGhutOAAAAOQmkiBkFb8/vnEAAADIPiRByCoOR3zjAAAAkH1IgpBV3G6rClzbhqdBNpvkcllxAAAAyE0kQcgqdrtVBluKTISCx1VV9P0BAADIZSRByDoej1RbK5WWho87ndZ4tD5BAAAAyG40S0VaCwSsSm5+v3Ufj9sd2yqOxyOVl3ftWgAAAGQ3kiCkLcOwev60LnntdFrb3WJZzbHbpbKyhE0PAAAAGYrtcEhLhiFVVET2/PH5rHHDSM28AAAAkPlIgpB2AgFrBcg0I88FxyorrTgAAACgs0iCkHa83sgVoNZMU2pstOIAAACAziIJQtrx++MbBwAAALRGEoS043DENw4AAABojSQIacfttqrAtW12GmSzSS6XFQcAAAB0FkkQ0o7dbpXBliIToeBxVRU9fwAAANA1JEFISx6PVFsrlZaGjzud1ngsfYIAAACA9tAsFUkRCFjV3Px+614etzv6So7HI5WXd/46AAAA4GhIgpBwhmH1/Wld9trptLa8RVvRsdulsrKETg8AAAA5hu1wSCjDkCoqIvv++HzWuGGkZl4AAADIXSRBSJhAwFoBMs3Ic8GxykorDgAAAEgWkiAkjNcbuQLUmmlKjY1WHAAAAJAsJEFIGL8/vnEAAABAPJAEIWEcjvjGAQAAAPFAEoSEcbutKnBtG54G2WySy2XFAQAAAMlCEoSEsdutMthSZCIUPK6qou8PAAAAkoskCAnl8Ui1tVJpafi402mNR+sTBAAAAMQbzVLRKYGAVc3N77fu5XG7o6/keDxSeXnnrwMAAAASgSQIMTMMq+9P67LXTqe15S3aio7dLpWVJXR6AAAAQEzYDoeYGIZUURHZ98fns8YNIzXzAgAAADorIUnQ6tWrZbPZ2n2sX79ekrRjx452z69duzYRU0I3BALWCpBpRp4LjlVWWnEAAABAukvIdrgJEybI36YD5ne/+12tWrVKZ599dtj4ypUrdcopp4SO+/fvn4gpoRu83sgVoNZMU2pstOLY8gYAAIB0l5AkKD8/XyUlJaHjw4cP6ze/+Y2uv/562drUSu7fv39YLNJPm3y223EAAABAKiXlnqBly5bpgw8+0NVXXx1x7tJLL9XAgQN1zjnnaNmyZVGf6+DBg2pubg57ILEcjvjGAQAAAKmUlCToySef1OTJk+V0OkNjffr00QMPPKAXXnhBv/vd73TOOedo6tSpUROhRYsWqaioKPRwuVyJnn7Oc7utKnBtG54G2WySy2XFAQAAAOnOZprt3e7evltvvVU//OEPjxpTV1enUaNGhY7ff/99DR06VEuWLNFll1121Guvuuoq1dfXy+v1dhhz8OBBHTx4MHTc3Nwsl8ulpqYmFRYWxvhK0FnB6nBSeIGEYGJE41MAAACkWnNzs4qKiqLmBp26J+imm27S7NmzjxpzwgknhB3X1NSof//+uvTSS6M+/7hx47RixYqjxhQUFKigoCDqcyG+PB4r0WmvT1BVFQkQAAAAMkenkqABAwZowIABMcebpqmamhpdddVV6tmzZ9T4jRs3ysGNJQkXCFiV3Px+6z4et9tqZhqNxyOVl3ftWgAAACBdJKQ6XNDLL7+s+vp6ff3rX4849/TTTys/P19nnnmmJMkwDD311FN64oknEjmlnGcY7a/mVFfHtppjt1MGGwAAAJktoUnQk08+qQkTJoTdI9Ta9773Pb333nvq0aOHRo0apeeff14VwRtPEHfB+3ra3gXm81nj3NcDAACAXNCpwgjpKNabn3JdICANG9Zx01ObzVoRqq9nexsAAAAyU6y5QVJKZCP1vN6OEyDJWh1qbLTiAAAAgGxGEpQj/P74xgEAAACZiiQoR8RadI/ifAAAAMh2JEE5wu227vkJNjdty2aTXC4rDgAAAMhmJEE5wm63ymBLkYlQ8LiqiqIIAAAAyH4kQTnE47HKYJeWho87nZTHBgAAQO5IaJ8gJFYgYFVz8/ute3nc7ugrOR6PVF7e+esAAACAbEESlKEMQ5o/P7zstdNpbXmLtqJjt0tlZQmdHgAAAJC22A6XgQxDqqiI7Pvj81njhpGaeQEAAACZgCQowwQC1gqQaUaeC45VVlpxAAAAACKRBGUYrzdyBag105QaG604AAAAAJFIgjKM3x/fOAAAACDXkARlGIcjvnEAAABAriEJyjBut1UFrm3D0yCbTXK5rDgAAAAAkUiCMozdbpXBliIToeBxVRV9fwAAAICOkARlII9Hqq2VSkvDx51OazxanyAAAAAgl9EsNcUCAauSm99v3cfjdse2iuPxSOXlXbsWAAAAyGUkQSlkGFbPn9Ylr51Oa7tbLKs5drtUVpaw6QEAAABZie1wKWIYUkVFZM8fn88aN4zUzAsAAADIdiRBKRAIWCtAphl5LjhWWWnFAQAAAIgvkqAU8HojV4BaM02psdGKAwAAABBfJEEp4PfHNw4AAABA7EiCUsDhiG8cAAAAgNiRBKWA221VgWvb7DTIZpNcLisOAAAAQHyRBKWA3W6VwZYiE6HgcVUVPX8AAACARCAJShGPR6qtlUpLw8edTms8lj5BAAAAADqPZqlxEghY1dz8futeHrc7+kqOxyOVl3f+OgAAAABdRxIUB4Zh9f1pXfba6bS2vEVb0bHbpbKyhE4PAAAAQCtsh+smw5AqKiL7/vh81rhhpGZeAAAAANpHEtQNgYC1AmSakeeCY5WVVhwAAACA9EAS1A1eb+QKUGumKTU2WnEAAAAA0gNJUDf4/fGNAwAAAJB4JEHd4HDENw4AAABA4pEEdYPbbVWBa9vwNMhmk1wuKw4AAABAeiAJ6ga73SqDLUUmQsHjqir6/gAAAADphCSomzweqbZWKi0NH3c6rfFofYIAAAAAJBfNUuPA45HKy60qcH6/dQ+Q280KEAAAAJCOSILixG6XyspSPQsAAAAA0bAdDgAAAEBOIQkCAAAAkFNIggAAAADkFJIgAAAAADmFJAgAAABATiEJAgAAAJBTSIIAAAAA5BSSIAAAAAA5hSQIAAAAQE4hCQIAAACQU0iCAAAAAOQUkiAAAAAAOYUkCAAAAEBOIQkCAAAAkFNIggAAAADkFJIgAAAAADmlR6on0F2maUqSmpubUzwTAAAAAKkUzAmCOUJHMj4J2rdvnyTJ5XKleCYAAAAA0sG+fftUVFTU4XmbGS1NSnMtLS3auXOn+vbtK5vNltK5NDc3y+VyqbGxUYWFhSmdSzbjfU4O3ufk4H1ODt7nxOM9Tg7e5+TgfU6ORLzPpmlq3759Gjx4sPLyOr7zJ+NXgvLy8uR0OlM9jTCFhYX8B5MEvM/JwfucHLzPycH7nHi8x8nB+5wcvM/JEe/3+WgrQEEURgAAAACQU0iCAAAAAOQUkqA4Kigo0B133KGCgoJUTyWr8T4nB+9zcvA+Jwfvc+LxHicH73Ny8D4nRyrf54wvjAAAAAAAncFKEAAAAICcQhIEAAAAIKeQBAEAAADIKSRBAAAAAHIKSRAAAACAnEIS1EU/+MEPNGHCBPXu3VvFxcXtxjQ0NOjiiy9W7969NXDgQH3729/WkSNHwmJWr16ts846SwUFBRoxYoQWL16c+MlnqNWrV8tms7X7WL9+vSRpx44d7Z5fu3ZtimefWYYNGxbxHt5zzz1hMZs2bZLb7dYxxxwjl8ule++9N0WzzUw7duzQNddco+HDh6tXr1468cQTdccdd+jQoUNhMXyeu++RRx7RsGHDdMwxx2jcuHF66623Uj2ljLZo0SJ9/vOfV9++fTVw4EBNnTpVW7ZsCYspKyuL+Nx+85vfTNGMM9Odd94Z8R6OGjUqdP7AgQOaO3eu+vfvrz59+uiyyy7T7t27UzjjzNTe/9/ZbDbNnTtXEp/lrnjttdd0ySWXaPDgwbLZbFq6dGnYedM0dfvtt8vhcKhXr16aNGmStm7dGhbz4Ycf6oorrlBhYaGKi4t1zTXX6OOPP47rPEmCuujQoUP6yle+om9961vtng8EArr44ot16NAhvfnmm3r66ae1ePFi3X777aGY+vp6XXzxxZo4caI2btyoyspKff3rX9dLL72UrJeRUSZMmCC/3x/2+PrXv67hw4fr7LPPDotduXJlWNzYsWNTNOvMdffdd4e9h9dff33oXHNzsy644AINHTpUGzZs0H333ac777xTjz/+eApnnFk2b96slpYWPfbYY/rb3/6mH/3oR3r00Uf13//93xGxfJ677vnnn9eCBQt0xx136M9//rNOP/10TZ48WXv27En11DLWq6++qrlz52rt2rVasWKFDh8+rAsuuED79+8Pi5szZ07Y55Y/lHTeKaecEvYevv7666FzN954o37729/qhRde0KuvvqqdO3fK4/GkcLaZaf369WHv8YoVKyRJX/nKV0IxfJY7Z//+/Tr99NP1yCOPtHv+3nvv1UMPPaRHH31U69at07HHHqvJkyfrwIEDoZgrrrhCf/vb37RixQq9+OKLeu2113TttdfGd6ImuqWmpsYsKiqKGP/9739v5uXlmbt27QqN/fSnPzULCwvNgwcPmqZpmt/5znfMU045Jey6yy+/3Jw8eXJC55wtDh06ZA4YMMC8++67Q2P19fWmJPPtt99O3cSywNChQ80f/ehHHZ7/yU9+Yh533HGhz7JpmuYtt9xijhw5Mgmzy1733nuvOXz48NAxn+fu+8IXvmDOnTs3dBwIBMzBgwebixYtSuGsssuePXtMSearr74aGjvvvPPM+fPnp25SWeCOO+4wTz/99HbP7d271+zZs6f5wgsvhMbq6upMSeaaNWuSNMPsNH/+fPPEE080W1paTNPks9xdksxf//rXoeOWlhazpKTEvO+++0Jje/fuNQsKCsznnnvONE3T/Pvf/25KMtevXx+K+cMf/mDabDbT5/PFbW6sBCXImjVrdOqpp2rQoEGhscmTJ6u5uVl/+9vfQjGTJk0Ku27y5Mlas2ZNUueaqZYtW6YPPvhAV199dcS5Sy+9VAMHDtQ555yjZcuWpWB2me+ee+5R//79deaZZ+q+++4L28q5Zs0anXvuucrPzw+NTZ48WVu2bNFHH32UiulmhaamJvXr1y9inM9z1xw6dEgbNmwI+zmbl5enSZMm8XM2jpqamiQp4rP7zDPP6Pjjj9eYMWO0cOFCffLJJ6mYXkbbunWrBg8erBNOOEFXXHGFGhoaJEkbNmzQ4cOHwz7bo0aN0pAhQ/hsd8OhQ4f0i1/8Ql/72tdks9lC43yW46e+vl67du0K++wWFRVp3Lhxoc/umjVrVFxcHLbLZ9KkScrLy9O6deviNpcecXsmhNm1a1dYAiQpdLxr166jxjQ3N+vTTz9Vr169kjPZDPXkk09q8uTJcjqdobE+ffrogQce0Be/+EXl5eXpV7/6laZOnaqlS5fq0ksvTeFsM8sNN9ygs846S/369dObb76phQsXyu/368EHH5RkfXaHDx8edk3rz/dxxx2X9Dlnum3btunhhx/W/fffHxrj89w9//rXvxQIBNr9Obt58+YUzSq7tLS0qLKyUl/84hc1ZsyY0PjMmTM1dOhQDR48WJs2bdItt9yiLVu2yDCMFM42s4wbN06LFy/WyJEj5ff7ddddd8ntduudd97Rrl27lJ+fH3FP8qBBg0K/Y6Dzli5dqr1792r27NmhMT7L8RX8fLb3c7n178cDBw4MO9+jRw/169cvrp9vkqBWbr31Vv3whz88akxdXV3YjYnovq687++//75eeuklLVmyJCzu+OOP14IFC0LHn//857Vz507dd999Of9LY2fe59bv4Wmnnab8/Hx94xvf0KJFi1RQUJDoqWa0rnyefT6fpkyZoq985SuaM2dOaJzPM9Ld3Llz9c4774TdqyIpbO/+qaeeKofDofPPP1/bt2/XiSeemOxpZqQLL7ww9O/TTjtN48aN09ChQ7VkyRL+SJogTz75pC688EINHjw4NMZnOXuRBLVy0003hWX/7TnhhBNieq6SkpKICkTBqi0lJSWhr20ruezevVuFhYU59QOuK+97TU2N+vfvH9MvguPGjQvd6JjLuvP5HjdunI4cOaIdO3Zo5MiRHX52pX9/vnNVZ9/nnTt3auLEiZowYUJMhSX4PMfu+OOPl91ub/ezmuuf03iYN29e6Ibl1ivy7Rk3bpwka8WTXxy7pri4WCeffLK2bdumL33pSzp06JD27t0bthrEZ7vr3nvvPa1cuTLqCg+f5e4Jfj53794th8MRGt+9e7fOOOOMUEzb4jVHjhzRhx9+GNfPN0lQKwMGDNCAAQPi8lzjx4/XD37wA+3Zsye0pLdixQoVFhbqc5/7XCjm97//fdh1K1as0Pjx4+Myh0zR2ffdNE3V1NToqquuUs+ePaPGb9y4Mew/tFzVnc/3xo0blZeXF/osjx8/Xv/zP/+jw4cPh/43WLFihUaOHJnzW+E68z77fD5NnDhRY8eOVU1NjfLyot+myec5dvn5+Ro7dqxWrVqlqVOnSrK2b61atUrz5s1L7eQymGmauv766/XrX/9aq1evjtga256NGzdKEp/dbvj444+1fft2XXnllRo7dqx69uypVatW6bLLLpMkbdmyRQ0NDTn3O0S81NTUaODAgbr44ouPGsdnuXuGDx+ukpISrVq1KpT0NDc3a926daGKy+PHj9fevXu1YcOGUDXUl19+WS0tLaEkNC7iVmIhx7z33nvm22+/bd51111mnz59zLffftt8++23zX379pmmaZpHjhwxx4wZY15wwQXmxo0bzeXLl5sDBgwwFy5cGHqOf/zjH2bv3r3Nb3/722ZdXZ35yCOPmHa73Vy+fHmqXlZGWLlypSnJrKurizi3ePFi89lnnzXr6urMuro68wc/+IGZl5dnPvXUUymYaWZ68803zR/96Efmxo0bze3bt5u/+MUvzAEDBphXXXVVKGbv3r3moEGDzCuvvNJ85513zF/+8pdm7969zcceeyyFM88s77//vjlixAjz/PPPN99//33T7/eHHkF8nrvvl7/8pVlQUGAuXrzY/Pvf/25ee+21ZnFxcVjlTnTOt771LbOoqMhcvXp12Of2k08+MU3TNLdt22befffd5p/+9Cezvr7e/M1vfmOecMIJ5rnnnpvimWeWm266yVy9erVZX19vvvHGG+akSZPM448/3tyzZ49pmqb5zW9+0xwyZIj58ssvm3/605/M8ePHm+PHj0/xrDNTIBAwhwwZYt5yyy1h43yWu2bfvn2h34slmQ8++KD59ttvm++9955pmqZ5zz33mMXFxeZvfvMbc9OmTWZ5ebk5fPhw89NPPw09x5QpU8wzzzzTXLdunfn666+bJ510kjljxoy4zpMkqItmzZplSop4vPLKK6GYHTt2mBdeeKHZq1cv8/jjjzdvuukm8/Dhw2HP88orr5hnnHGGmZ+fb55wwglmTU1Ncl9IBpoxY4Y5YcKEds8tXrzYHD16tNm7d2+zsLDQ/MIXvhBWQhTRbdiwwRw3bpxZVFRkHnPMMebo0aPN//3f/zUPHDgQFveXv/zFPOecc8yCggKztLTUvOeee1I048xUU1PT7s+Q1n+b4vMcHw8//LA5ZMgQMz8/3/zCF75grl27NtVTymgdfW6D///V0NBgnnvuuWa/fv3MgoICc8SIEea3v/1ts6mpKbUTzzCXX3656XA4zPz8fLO0tNS8/PLLzW3btoXOf/rpp+Z1111nHnfccWbv3r3NadOmhf0RBbF76aWXTEnmli1bwsb5LHfNK6+80u7PiFmzZpmmaZXJ/u53v2sOGjTILCgoMM8///yI9/6DDz4wZ8yYYfbp08csLCw0r7766tBCQ7zYTNM047euBAAAAADpjT5BAAAAAHIKSRAAAACAnEISBAAAACCnkAQBAAAAyCkkQQAAAAByCkkQAAAAgJxCEgQAAAAgp5AEAQAAAMgpJEEAAAAAcgpJEAAAAICcQhIEAAAAIKf8P1es0ogouU38AAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate model_3 evaluation metrics\n",
        "mae_3 = mae(y_test, y_preds_3)\n",
        "mse_3 = mse(y_test, y_preds_3)\n",
        "mae_3, mse_3"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qwWPf5w8jJGE",
        "outputId": "de9cdf35-4141-4c18-ae32-2c728b1a39c0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(<tf.Tensor: shape=(), dtype=float32, numpy=68.69386>,\n",
              " <tf.Tensor: shape=(), dtype=float32, numpy=4805.2993>)"
            ]
          },
          "metadata": {},
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** You want to start with small experiments (small models) and make sure they work and then increase their scale when necessary."
      ],
      "metadata": {
        "id": "2bd1dvZ8k7DN"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Comparing and tracking your TensorFlow modelling experiments\n",
        "\n",
        "We've run a few experiments, let's compare the results."
      ],
      "metadata": {
        "id": "Lgy9nfGakS5J"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Let's compare our model's results using a pandas DataFrame\n",
        "import pandas as pd\n",
        "\n",
        "model_results = [[\"model_1\", mae_1.numpy(), mse_1.numpy()],\n",
        "                 [\"model_2\", mae_2.numpy(), mse_2.numpy()],\n",
        "                 [\"model_3\", mae_3.numpy(), mse_3.numpy()]]\n",
        "\n",
        "all_results = pd.DataFrame(model_results, columns=[\"model\", \"mae\", \"mse\"])\n",
        "all_results"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 143
        },
        "id": "ALndiCgJlSUh",
        "outputId": "41658556-5074-441c-99bd-ebecf5153f54"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     model        mae          mse\n",
              "0  model_1  21.338467   458.786530\n",
              "1  model_2  11.590208   142.928452\n",
              "2  model_3  68.693863  4805.299316"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-830a2d73-c1ac-4234-98ac-1dabe7667bb1\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>model</th>\n",
              "      <th>mae</th>\n",
              "      <th>mse</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>model_1</td>\n",
              "      <td>21.338467</td>\n",
              "      <td>458.786530</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>model_2</td>\n",
              "      <td>11.590208</td>\n",
              "      <td>142.928452</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>model_3</td>\n",
              "      <td>68.693863</td>\n",
              "      <td>4805.299316</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-830a2d73-c1ac-4234-98ac-1dabe7667bb1')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-830a2d73-c1ac-4234-98ac-1dabe7667bb1 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-830a2d73-c1ac-4234-98ac-1dabe7667bb1');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-2144b306-0fc1-4bf6-869f-9d775556062f\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-2144b306-0fc1-4bf6-869f-9d775556062f')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-2144b306-0fc1-4bf6-869f-9d775556062f button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_145db622-fb28-4643-8ef9-c8635b47c542\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('all_results')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_145db622-fb28-4643-8ef9-c8635b47c542 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('all_results');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "all_results",
              "summary": "{\n  \"name\": \"all_results\",\n  \"rows\": 3,\n  \"fields\": [\n    {\n      \"column\": \"model\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          \"model_1\",\n          \"model_2\",\n          \"model_3\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"mae\",\n      \"properties\": {\n        \"dtype\": \"float32\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          21.33846664428711,\n          11.590208053588867,\n          68.69386291503906\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"mse\",\n      \"properties\": {\n        \"dtype\": \"float32\",\n        \"num_unique_values\": 3,\n        \"samples\": [\n          458.7865295410156,\n          142.92845153808594,\n          4805.29931640625\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 55
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Looks like `model_2` performed the best..."
      ],
      "metadata": {
        "id": "KxGBGzLumgJm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "98X_tVY2liWI",
        "outputId": "46724b80-a4a5-4707-cd60-29bf798db9cc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 10)                20        \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 31 (124.00 Byte)\n",
            "Trainable params: 31 (124.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Note:** One of your main goals should be to minimize the time between your experiments. The more experiments you do, the more things you'll figure out which don't work and in turn, you will get closer to figuring out what does work. Remember the machine learning practioner's motto: \"experiment, experiment, experiment\"."
      ],
      "metadata": {
        "id": "y1rTdnPRl9KJ"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Tracking your experiments\n",
        "\n",
        "One really good habit in machine larning modelling is to track the results of your experiments.\n",
        "\n",
        "And when doing so, it can be tedious if you're running lots of experiments.\n",
        "\n",
        "Luckily, there are tools to help us!\n",
        "\n",
        "**Resource:** As you build more models, you'll want to look into using:\n",
        "* TensorBoard - a component of the TensorFlow library to help track modelling experiments (we'll see this one later). https://www.tensorflow.org/tensorboard\n",
        "* Weights & Biases - a tool for tracking all kinds of machine learning experiments (plugs straight into TensorBoard). https://wandb.ai/site"
      ],
      "metadata": {
        "id": "CWYKPVktmnez"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## How to save a TensorFlow model\n",
        "\n",
        "Saving our models allows us to use them outside of Google Colab (or wherever they are trained) such as in a web application or a mobile app.\n",
        "\n",
        "There are two main formats we can save our model's too:\n",
        "1. The SavedModel format\n",
        "2. The HDF5 format\n",
        "\n",
        "https://www.tensorflow.org/tutorials/keras/save_and_load"
      ],
      "metadata": {
        "id": "si8bauOUnErH"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Save model using the SavedFormat format\n",
        "model_2.save(\"best_model_SavedModel_format\")"
      ],
      "metadata": {
        "id": "ChqrMjBmqwex"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Save model using the SavedModel format\n",
        "model_2.save(\"best_model_HDF5_format.h5\")"
      ],
      "metadata": {
        "id": "2lpEqB62q6Hj",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "16b9dbd2-a1f6-40f7-e18d-884799c76144"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3103: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## How to load and use a saved TensorFlow model"
      ],
      "metadata": {
        "id": "H6Gt-3_TsGsS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Load in the Saved model format\n",
        "loaded_SavedModel_format = tf.keras.models.load_model(\"best_model_SavedModel_format\")\n",
        "loaded_SavedModel_format.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aDRfTi3rswyc",
        "outputId": "037ccb8b-34f7-461f-f9b4-ff53ecb6045a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 10)                20        \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 31 (124.00 Byte)\n",
            "Trainable params: 31 (124.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "15-Ey2ZgtIL3",
        "outputId": "4867a5b7-b8a3-46aa-de63-389a91a561b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 10)                20        \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 31 (124.00 Byte)\n",
            "Trainable params: 31 (124.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Compare model_2 predictions with SavedModel format model predictions\n",
        "model_2_preds = model_2.predict(X_test)\n",
        "loaded_SavedModel_format_preds = loaded_SavedModel_format.predict(X_test)\n",
        "model_2_preds == loaded_SavedModel_format_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XZPmXQTmtO2S",
        "outputId": "24659206-33ed-4bcf-b3b5-aa9592c58aeb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 51ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True]])"
            ]
          },
          "metadata": {},
          "execution_count": 61
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load in a file using the .h5 format\n",
        "loaded_h5_model = tf.keras.models.load_model(\"/content/best_model_HDF5_format.h5\")\n",
        "loaded_h5_model.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "alEMTHrlutYW",
        "outputId": "db24e928-1e55-4913-ae6b-76f30a13313e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 10)                20        \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 31 (124.00 Byte)\n",
            "Trainable params: 31 (124.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eQ8-XqqFvep9",
        "outputId": "4c5d7957-a5b3-40b6-d64a-4fa625505270"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential_5\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense_7 (Dense)             (None, 10)                20        \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 1)                 11        \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 31 (124.00 Byte)\n",
            "Trainable params: 31 (124.00 Byte)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check to see if loaded .h5 model predictions match model_2\n",
        "model_2_preds == model_2.predict(X_test)\n",
        "loaded_h5_model_preds = loaded_h5_model.predict(X_test)\n",
        "model_2_preds == loaded_h5_model_preds"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ln6qJ30wvhEO",
        "outputId": "54bff728-a10d-46b4-e930-2e2b7ff7ec18"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 19ms/step\n",
            "1/1 [==============================] - 0s 55ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True],\n",
              "       [ True]])"
            ]
          },
          "metadata": {},
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## How to save and download files from Google Colab\n",
        "\n",
        "If you want to download your files from Google Colab:\n",
        "1. You can go to the \"files\" tab and right click on the file you're after and click \"download\".\n",
        "2. Use code (see the cell below)\n",
        "3. Save it to Google Drive by connecting Google Drive and copying it to there."
      ],
      "metadata": {
        "id": "VLMQuS-4vuuR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Download a file from Google Colab\n",
        "from google.colab import files\n",
        "# files.download(\"/content/best_model_HDF5_format.h5\")"
      ],
      "metadata": {
        "id": "f93gEpsaxXaP"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Save a file from Google Colab to Google Drive (requires mounting Google Drive)\n",
        "!cp /content/best_model_HDF5_format.h5 \"/content/drive/MyDrive/Udemy/TensorFlow\""
      ],
      "metadata": {
        "id": "WAUqMDtsx3u7",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ea965025-af8a-4f54-8e5a-dc448e275c2a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cp: cannot create regular file '/content/drive/MyDrive/Udemy/TensorFlow': No such file or directory\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!ls \"/content/drive/MyDrive/Udemy/TensorFlow\""
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "j8FM0mmuyzyy",
        "outputId": "6e1cd1ad-4bca-41bf-9439-aea4dfb2d3d9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "ls: cannot access '/content/drive/MyDrive/Udemy/TensorFlow': No such file or directory\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Putting together what we've learned"
      ],
      "metadata": {
        "id": "aafnnte9y3g4"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Part 1 (preparing a dataset)\n",
        "\n",
        "https://www.kaggle.com/datasets/mirichoi0218/insurance"
      ],
      "metadata": {
        "id": "Y17K8gm3zd8b"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Import required libraries\n",
        "import tensorflow as tf\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "6T8bOegpzgHZ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Read in the insurance dataset\n",
        "insurance = pd.read_csv(\"https://raw.githubusercontent.com/stedy/Machine-Learning-with-R-datasets/master/insurance.csv\")\n",
        "insurance"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "8-2CLwo90jnP",
        "outputId": "68eaeef9-a9c2-47df-84e6-258b9e32dd7f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      age     sex     bmi  children smoker     region      charges\n",
              "0      19  female  27.900         0    yes  southwest  16884.92400\n",
              "1      18    male  33.770         1     no  southeast   1725.55230\n",
              "2      28    male  33.000         3     no  southeast   4449.46200\n",
              "3      33    male  22.705         0     no  northwest  21984.47061\n",
              "4      32    male  28.880         0     no  northwest   3866.85520\n",
              "...   ...     ...     ...       ...    ...        ...          ...\n",
              "1333   50    male  30.970         3     no  northwest  10600.54830\n",
              "1334   18  female  31.920         0     no  northeast   2205.98080\n",
              "1335   18  female  36.850         0     no  southeast   1629.83350\n",
              "1336   21  female  25.800         0     no  southwest   2007.94500\n",
              "1337   61  female  29.070         0    yes  northwest  29141.36030\n",
              "\n",
              "[1338 rows x 7 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-45b9de29-9722-4076-b687-f1f2147eb409\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>sex</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>smoker</th>\n",
              "      <th>region</th>\n",
              "      <th>charges</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>female</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southwest</td>\n",
              "      <td>16884.92400</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>male</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1725.55230</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>male</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>4449.46200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>male</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>21984.47061</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>male</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>3866.85520</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1333</th>\n",
              "      <td>50</td>\n",
              "      <td>male</td>\n",
              "      <td>30.970</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>10600.54830</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1334</th>\n",
              "      <td>18</td>\n",
              "      <td>female</td>\n",
              "      <td>31.920</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northeast</td>\n",
              "      <td>2205.98080</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1335</th>\n",
              "      <td>18</td>\n",
              "      <td>female</td>\n",
              "      <td>36.850</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1629.83350</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1336</th>\n",
              "      <td>21</td>\n",
              "      <td>female</td>\n",
              "      <td>25.800</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>southwest</td>\n",
              "      <td>2007.94500</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1337</th>\n",
              "      <td>61</td>\n",
              "      <td>female</td>\n",
              "      <td>29.070</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>northwest</td>\n",
              "      <td>29141.36030</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1338 rows × 7 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-45b9de29-9722-4076-b687-f1f2147eb409')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-45b9de29-9722-4076-b687-f1f2147eb409 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-45b9de29-9722-4076-b687-f1f2147eb409');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-a0561415-f67c-461d-81de-b3446aad0b03\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-a0561415-f67c-461d-81de-b3446aad0b03')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-a0561415-f67c-461d-81de-b3446aad0b03 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_8766ee4f-aa34-444e-a80c-43b5d5c4f1d7\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('insurance')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_8766ee4f-aa34-444e-a80c-43b5d5c4f1d7 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('insurance');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "insurance",
              "summary": "{\n  \"name\": \"insurance\",\n  \"rows\": 1338,\n  \"fields\": [\n    {\n      \"column\": \"age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 14,\n        \"min\": 18,\n        \"max\": 64,\n        \"num_unique_values\": 47,\n        \"samples\": [\n          21,\n          45,\n          36\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"male\",\n          \"female\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"bmi\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 6.098186911679014,\n        \"min\": 15.96,\n        \"max\": 53.13,\n        \"num_unique_values\": 548,\n        \"samples\": [\n          23.18,\n          26.885\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"children\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 0,\n        \"max\": 5,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"no\",\n          \"yes\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"southeast\",\n          \"northeast\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"charges\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 12110.011236694001,\n        \"min\": 1121.8739,\n        \"max\": 63770.42801,\n        \"num_unique_values\": 1337,\n        \"samples\": [\n          8688.85885,\n          5708.867\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check the datatypes of the data in each columns\n",
        "insurance.dtypes"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "471AFMY50pj8",
        "outputId": "c3581971-5f7f-4efb-b2aa-19d1030989b8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "age           int64\n",
              "sex          object\n",
              "bmi         float64\n",
              "children      int64\n",
              "smoker       object\n",
              "region       object\n",
              "charges     float64\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# One hot encode categorical variables\n",
        "# https://pandas.pydata.org/docs/reference/api/pandas.get_dummies.html\n",
        "insurance_one_hot = pd.get_dummies(insurance)\n",
        "insurance_one_hot.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "5wU31tnP1VM-",
        "outputId": "d08cda4f-8eaf-4c86-bc99-29607f558c22"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age     bmi  children      charges  sex_female  sex_male  smoker_no  \\\n",
              "0   19  27.900         0  16884.92400           1         0          0   \n",
              "1   18  33.770         1   1725.55230           0         1          1   \n",
              "2   28  33.000         3   4449.46200           0         1          1   \n",
              "3   33  22.705         0  21984.47061           0         1          1   \n",
              "4   32  28.880         0   3866.85520           0         1          1   \n",
              "\n",
              "   smoker_yes  region_northeast  region_northwest  region_southeast  \\\n",
              "0           1                 0                 0                 0   \n",
              "1           0                 0                 0                 1   \n",
              "2           0                 0                 0                 1   \n",
              "3           0                 0                 1                 0   \n",
              "4           0                 0                 1                 0   \n",
              "\n",
              "   region_southwest  \n",
              "0                 1  \n",
              "1                 0  \n",
              "2                 0  \n",
              "3                 0  \n",
              "4                 0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9479dd80-a6f8-4eef-8893-650b5d7bfc97\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>charges</th>\n",
              "      <th>sex_female</th>\n",
              "      <th>sex_male</th>\n",
              "      <th>smoker_no</th>\n",
              "      <th>smoker_yes</th>\n",
              "      <th>region_northeast</th>\n",
              "      <th>region_northwest</th>\n",
              "      <th>region_southeast</th>\n",
              "      <th>region_southwest</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>16884.92400</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>1725.55230</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>4449.46200</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>21984.47061</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>3866.85520</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9479dd80-a6f8-4eef-8893-650b5d7bfc97')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-9479dd80-a6f8-4eef-8893-650b5d7bfc97 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-9479dd80-a6f8-4eef-8893-650b5d7bfc97');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-59317863-e40e-4ffd-baac-c09605993855\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-59317863-e40e-4ffd-baac-c09605993855')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-59317863-e40e-4ffd-baac-c09605993855 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "insurance_one_hot",
              "summary": "{\n  \"name\": \"insurance_one_hot\",\n  \"rows\": 1338,\n  \"fields\": [\n    {\n      \"column\": \"age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 14,\n        \"min\": 18,\n        \"max\": 64,\n        \"num_unique_values\": 47,\n        \"samples\": [\n          21,\n          45,\n          36\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"bmi\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 6.098186911679014,\n        \"min\": 15.96,\n        \"max\": 53.13,\n        \"num_unique_values\": 548,\n        \"samples\": [\n          23.18,\n          26.885,\n          29.26\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"children\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 0,\n        \"max\": 5,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          0,\n          1,\n          4\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"charges\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 12110.011236694001,\n        \"min\": 1121.8739,\n        \"max\": 63770.42801,\n        \"num_unique_values\": 1337,\n        \"samples\": [\n          8688.85885,\n          5708.867,\n          11436.73815\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex_female\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex_male\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker_no\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker_yes\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_northeast\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_northwest\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_southeast\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_southwest\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 71
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "### Part 2 (building a regression model)"
      ],
      "metadata": {
        "id": "tCFskE_72Uc7"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Create X & y values (features and labels)\n",
        "X = insurance_one_hot.drop(\"charges\", axis=1)\n",
        "y = insurance_one_hot[\"charges\"]"
      ],
      "metadata": {
        "id": "WWSuz4Sh4Bwh"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# View X\n",
        "X.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "GeoevHt34EdZ",
        "outputId": "ca15de9c-936a-4f4b-ac93-63e2444f49e5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "   age     bmi  children  sex_female  sex_male  smoker_no  smoker_yes  \\\n",
              "0   19  27.900         0           1         0          0           1   \n",
              "1   18  33.770         1           0         1          1           0   \n",
              "2   28  33.000         3           0         1          1           0   \n",
              "3   33  22.705         0           0         1          1           0   \n",
              "4   32  28.880         0           0         1          1           0   \n",
              "\n",
              "   region_northeast  region_northwest  region_southeast  region_southwest  \n",
              "0                 0                 0                 0                 1  \n",
              "1                 0                 0                 1                 0  \n",
              "2                 0                 0                 1                 0  \n",
              "3                 0                 1                 0                 0  \n",
              "4                 0                 1                 0                 0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-b16b1a4c-4f81-423a-9df9-6690ec7b64d5\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>sex_female</th>\n",
              "      <th>sex_male</th>\n",
              "      <th>smoker_no</th>\n",
              "      <th>smoker_yes</th>\n",
              "      <th>region_northeast</th>\n",
              "      <th>region_northwest</th>\n",
              "      <th>region_southeast</th>\n",
              "      <th>region_southwest</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b16b1a4c-4f81-423a-9df9-6690ec7b64d5')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-b16b1a4c-4f81-423a-9df9-6690ec7b64d5 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-b16b1a4c-4f81-423a-9df9-6690ec7b64d5');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-8649172d-8d9e-4c0a-8589-cbdbe298aae7\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-8649172d-8d9e-4c0a-8589-cbdbe298aae7')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-8649172d-8d9e-4c0a-8589-cbdbe298aae7 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "X",
              "summary": "{\n  \"name\": \"X\",\n  \"rows\": 1338,\n  \"fields\": [\n    {\n      \"column\": \"age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 14,\n        \"min\": 18,\n        \"max\": 64,\n        \"num_unique_values\": 47,\n        \"samples\": [\n          21,\n          45,\n          36\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"bmi\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 6.098186911679014,\n        \"min\": 15.96,\n        \"max\": 53.13,\n        \"num_unique_values\": 548,\n        \"samples\": [\n          23.18,\n          26.885,\n          29.26\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"children\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 0,\n        \"max\": 5,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          0,\n          1,\n          4\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex_female\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex_male\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker_no\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker_yes\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_northeast\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_northwest\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_southeast\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region_southwest\",\n      \"properties\": {\n        \"dtype\": \"uint8\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# View y\n",
        "y.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YWNrFdds4QkB",
        "outputId": "778a355d-a962-45a7-ba2b-4914dd154650"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0    16884.92400\n",
              "1     1725.55230\n",
              "2     4449.46200\n",
              "3    21984.47061\n",
              "4     3866.85520\n",
              "Name: charges, dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create training and test sets\n",
        "# https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "len(X), len(X_train), len(X_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H9B8xUpM4Rer",
        "outputId": "479456dc-d8cf-483a-bfb3-e3f6080f2f13"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1338, 1070, 268)"
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Build a neural network (sort of like model_2 above)\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create a model\n",
        "insurance_model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "insurance_model.compile(loss=tf.keras.losses.mae,\n",
        "                        optimizer=tf.keras.optimizers.SGD(),\n",
        "                        metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "insurance_model.fit(X_train, y_train, epochs=100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GJr0qmJ05sF2",
        "outputId": "f2449c67-d962-4cfc-ac0d-ea0a3727a6f9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8557.7471 - mae: 8557.7471\n",
            "Epoch 2/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7884.2017 - mae: 7884.2017\n",
            "Epoch 3/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7530.7700 - mae: 7530.7700\n",
            "Epoch 4/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7695.0903 - mae: 7695.0903\n",
            "Epoch 5/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7739.4712 - mae: 7739.4712\n",
            "Epoch 6/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7659.2773 - mae: 7659.2773\n",
            "Epoch 7/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7530.8003 - mae: 7530.8003\n",
            "Epoch 8/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7721.5786 - mae: 7721.5786\n",
            "Epoch 9/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7555.8740 - mae: 7555.8740\n",
            "Epoch 10/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7675.9619 - mae: 7675.9619\n",
            "Epoch 11/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7438.6821 - mae: 7438.6821\n",
            "Epoch 12/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7704.6587 - mae: 7704.6587\n",
            "Epoch 13/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7677.3174 - mae: 7677.3174\n",
            "Epoch 14/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7751.1743 - mae: 7751.1743\n",
            "Epoch 15/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7400.6387 - mae: 7400.6387\n",
            "Epoch 16/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7777.2773 - mae: 7777.2773\n",
            "Epoch 17/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7513.5596 - mae: 7513.5596\n",
            "Epoch 18/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7845.3423 - mae: 7845.3423\n",
            "Epoch 19/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7700.2695 - mae: 7700.2695\n",
            "Epoch 20/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7954.3599 - mae: 7954.3599\n",
            "Epoch 21/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7447.3218 - mae: 7447.3218\n",
            "Epoch 22/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7882.7783 - mae: 7882.7783\n",
            "Epoch 23/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7639.7935 - mae: 7639.7935\n",
            "Epoch 24/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7560.5103 - mae: 7560.5103\n",
            "Epoch 25/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7688.8130 - mae: 7688.8130\n",
            "Epoch 26/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7619.5112 - mae: 7619.5112\n",
            "Epoch 27/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7552.0981 - mae: 7552.0981\n",
            "Epoch 28/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7392.9004 - mae: 7392.9004\n",
            "Epoch 29/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7655.9092 - mae: 7655.9092\n",
            "Epoch 30/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7536.1685 - mae: 7536.1685\n",
            "Epoch 31/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7696.2417 - mae: 7696.2417\n",
            "Epoch 32/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7472.5703 - mae: 7472.5703\n",
            "Epoch 33/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7435.9302 - mae: 7435.9302\n",
            "Epoch 34/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7396.2197 - mae: 7396.2197\n",
            "Epoch 35/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7615.0215 - mae: 7615.0215\n",
            "Epoch 36/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7572.1431 - mae: 7572.1431\n",
            "Epoch 37/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7561.5996 - mae: 7561.5996\n",
            "Epoch 38/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7401.7544 - mae: 7401.7544\n",
            "Epoch 39/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7540.3096 - mae: 7540.3096\n",
            "Epoch 40/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7385.9395 - mae: 7385.9395\n",
            "Epoch 41/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7714.2988 - mae: 7714.2988\n",
            "Epoch 42/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7510.6802 - mae: 7510.6802\n",
            "Epoch 43/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7546.3013 - mae: 7546.3013\n",
            "Epoch 44/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7403.8022 - mae: 7403.8022\n",
            "Epoch 45/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7490.1284 - mae: 7490.1284\n",
            "Epoch 46/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7389.0894 - mae: 7389.0894\n",
            "Epoch 47/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7428.1279 - mae: 7428.1279\n",
            "Epoch 48/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7453.7300 - mae: 7453.7300\n",
            "Epoch 49/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7552.1724 - mae: 7552.1724\n",
            "Epoch 50/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7478.1240 - mae: 7478.1240\n",
            "Epoch 51/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7428.1792 - mae: 7428.1792\n",
            "Epoch 52/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7379.3149 - mae: 7379.3149\n",
            "Epoch 53/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7441.8169 - mae: 7441.8169\n",
            "Epoch 54/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7378.3398 - mae: 7378.3398\n",
            "Epoch 55/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7289.8853 - mae: 7289.8853\n",
            "Epoch 56/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7380.1958 - mae: 7380.1958\n",
            "Epoch 57/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7497.9360 - mae: 7497.9360\n",
            "Epoch 58/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7569.2212 - mae: 7569.2212\n",
            "Epoch 59/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7523.9214 - mae: 7523.9214\n",
            "Epoch 60/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7467.7339 - mae: 7467.7339\n",
            "Epoch 61/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7506.4424 - mae: 7506.4424\n",
            "Epoch 62/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7491.2319 - mae: 7491.2319\n",
            "Epoch 63/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7324.6304 - mae: 7324.6304\n",
            "Epoch 64/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7354.2700 - mae: 7354.2700\n",
            "Epoch 65/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7489.5479 - mae: 7489.5479\n",
            "Epoch 66/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7295.0464 - mae: 7295.0464\n",
            "Epoch 67/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7466.4287 - mae: 7466.4287\n",
            "Epoch 68/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7473.6440 - mae: 7473.6440\n",
            "Epoch 69/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7598.7290 - mae: 7598.7290\n",
            "Epoch 70/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7562.6562 - mae: 7562.6562\n",
            "Epoch 71/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7165.8203 - mae: 7165.8203\n",
            "Epoch 72/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7119.5796 - mae: 7119.5796\n",
            "Epoch 73/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7518.2739 - mae: 7518.2739\n",
            "Epoch 74/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7506.3354 - mae: 7506.3354\n",
            "Epoch 75/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7329.9526 - mae: 7329.9526\n",
            "Epoch 76/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7187.5229 - mae: 7187.5229\n",
            "Epoch 77/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7516.7607 - mae: 7516.7607\n",
            "Epoch 78/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7166.8169 - mae: 7166.8169\n",
            "Epoch 79/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7488.3872 - mae: 7488.3872\n",
            "Epoch 80/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7254.8784 - mae: 7254.8784\n",
            "Epoch 81/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7273.3218 - mae: 7273.3218\n",
            "Epoch 82/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7071.2979 - mae: 7071.2979\n",
            "Epoch 83/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7471.9946 - mae: 7471.9946\n",
            "Epoch 84/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7293.3213 - mae: 7293.3213\n",
            "Epoch 85/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7548.4292 - mae: 7548.4292\n",
            "Epoch 86/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7398.1416 - mae: 7398.1416\n",
            "Epoch 87/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7288.2339 - mae: 7288.2339\n",
            "Epoch 88/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7417.7075 - mae: 7417.7075\n",
            "Epoch 89/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7493.1387 - mae: 7493.1387\n",
            "Epoch 90/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6976.1870 - mae: 6976.1870\n",
            "Epoch 91/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7208.9419 - mae: 7208.9419\n",
            "Epoch 92/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7272.5479 - mae: 7272.5479\n",
            "Epoch 93/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7492.6318 - mae: 7492.6318\n",
            "Epoch 94/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7156.4038 - mae: 7156.4038\n",
            "Epoch 95/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7291.1719 - mae: 7291.1719\n",
            "Epoch 96/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7106.3853 - mae: 7106.3853\n",
            "Epoch 97/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7095.1855 - mae: 7095.1855\n",
            "Epoch 98/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7309.4165 - mae: 7309.4165\n",
            "Epoch 99/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 7346.0093 - mae: 7346.0093\n",
            "Epoch 100/100\n",
            "34/34 [==============================] - 0s 1ms/step - loss: 6947.7500 - mae: 6947.7500\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba275b66d70>"
            ]
          },
          "metadata": {},
          "execution_count": 76
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Check the results of the insurance model on the test data\n",
        "insurance_model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DJ177R-P6oz0",
        "outputId": "b0e0f66a-882e-47e4-8727-f712e0be42aa"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 7668.9380 - mae: 7668.9380\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[7668.93798828125, 7668.93798828125]"
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_train.median(), y_train.mean()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LAF0XQRP7PtP",
        "outputId": "2bd4b503-bb05-4597-8efc-7c1b40b6edd2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(9575.4421, 13346.089736364485)"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Right now it looks like our model isn't performing too well... let's try to improve it!\n",
        "\n",
        "To (try) improve our model, we'll run 2 experiments:\n",
        "1. Add an extra layer with more hidden units and use the Adam optimizer\n",
        "2. Same as above but train for longer (200 epochs)\n",
        "3. (Insert your own experiment here)"
      ],
      "metadata": {
        "id": "xmCSfFia7UNu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create the model\n",
        "insurance_model_2 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(100),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "insurance_model_2.compile(loss=tf.keras.losses.mae,\n",
        "                          optimizer=tf.keras.optimizers.Adam(),\n",
        "                          metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "insurance_model_2.fit(X_train, y_train, epochs=100, verbose=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JFt1t-RT7yT4",
        "outputId": "3f308dba-da60-4976-d4e1-a0baeaa5489b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "34/34 [==============================] - 1s 2ms/step - loss: 13296.4404 - mae: 13296.4404\n",
            "Epoch 2/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13131.1113 - mae: 13131.1113\n",
            "Epoch 3/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12779.2910 - mae: 12779.2910\n",
            "Epoch 4/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12083.0059 - mae: 12083.0059\n",
            "Epoch 5/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 10924.8281 - mae: 10924.8281\n",
            "Epoch 6/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 9472.2070 - mae: 9472.2070\n",
            "Epoch 7/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8168.9756 - mae: 8168.9756\n",
            "Epoch 8/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7551.2397 - mae: 7551.2397\n",
            "Epoch 9/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7450.7690 - mae: 7450.7690\n",
            "Epoch 10/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7430.3374 - mae: 7430.3374\n",
            "Epoch 11/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7411.3330 - mae: 7411.3330\n",
            "Epoch 12/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7389.1611 - mae: 7389.1611\n",
            "Epoch 13/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7368.6328 - mae: 7368.6328\n",
            "Epoch 14/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7346.1509 - mae: 7346.1509\n",
            "Epoch 15/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7326.9839 - mae: 7326.9839\n",
            "Epoch 16/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7304.4077 - mae: 7304.4077\n",
            "Epoch 17/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7284.1289 - mae: 7284.1289\n",
            "Epoch 18/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7260.5801 - mae: 7260.5801\n",
            "Epoch 19/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7238.1777 - mae: 7238.1777\n",
            "Epoch 20/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7214.9092 - mae: 7214.9092\n",
            "Epoch 21/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7195.5366 - mae: 7195.5366\n",
            "Epoch 22/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7168.9897 - mae: 7168.9897\n",
            "Epoch 23/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7144.1240 - mae: 7144.1240\n",
            "Epoch 24/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7119.5063 - mae: 7119.5063\n",
            "Epoch 25/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7101.4878 - mae: 7101.4878\n",
            "Epoch 26/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7069.9995 - mae: 7069.9995\n",
            "Epoch 27/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7041.5601 - mae: 7041.5601\n",
            "Epoch 28/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7013.8677 - mae: 7013.8677\n",
            "Epoch 29/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6985.8750 - mae: 6985.8750\n",
            "Epoch 30/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6959.2812 - mae: 6959.2812\n",
            "Epoch 31/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6928.1230 - mae: 6928.1230\n",
            "Epoch 32/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6901.1631 - mae: 6901.1631\n",
            "Epoch 33/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6870.1104 - mae: 6870.1104\n",
            "Epoch 34/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6838.6064 - mae: 6838.6064\n",
            "Epoch 35/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6805.8809 - mae: 6805.8809\n",
            "Epoch 36/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6772.0493 - mae: 6772.0493\n",
            "Epoch 37/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6736.7446 - mae: 6736.7446\n",
            "Epoch 38/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6704.7808 - mae: 6704.7808\n",
            "Epoch 39/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6665.8774 - mae: 6665.8774\n",
            "Epoch 40/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6630.5928 - mae: 6630.5928\n",
            "Epoch 41/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6597.6802 - mae: 6597.6802\n",
            "Epoch 42/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6566.7095 - mae: 6566.7095\n",
            "Epoch 43/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6536.0830 - mae: 6536.0830\n",
            "Epoch 44/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6510.3223 - mae: 6510.3223\n",
            "Epoch 45/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6492.3970 - mae: 6492.3970\n",
            "Epoch 46/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6473.5771 - mae: 6473.5771\n",
            "Epoch 47/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6456.6401 - mae: 6456.6401\n",
            "Epoch 48/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6441.0444 - mae: 6441.0444\n",
            "Epoch 49/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6427.6709 - mae: 6427.6709\n",
            "Epoch 50/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6413.5146 - mae: 6413.5146\n",
            "Epoch 51/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6399.5640 - mae: 6399.5640\n",
            "Epoch 52/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6387.8242 - mae: 6387.8242\n",
            "Epoch 53/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6373.8262 - mae: 6373.8262\n",
            "Epoch 54/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6359.4072 - mae: 6359.4072\n",
            "Epoch 55/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6345.8765 - mae: 6345.8765\n",
            "Epoch 56/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6330.9165 - mae: 6330.9165\n",
            "Epoch 57/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6318.4180 - mae: 6318.4180\n",
            "Epoch 58/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6304.2017 - mae: 6304.2017\n",
            "Epoch 59/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6288.6108 - mae: 6288.6108\n",
            "Epoch 60/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6277.6982 - mae: 6277.6982\n",
            "Epoch 61/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6257.7251 - mae: 6257.7251\n",
            "Epoch 62/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6245.1426 - mae: 6245.1426\n",
            "Epoch 63/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6226.2920 - mae: 6226.2920\n",
            "Epoch 64/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6209.9126 - mae: 6209.9126\n",
            "Epoch 65/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6192.1074 - mae: 6192.1074\n",
            "Epoch 66/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6174.5869 - mae: 6174.5869\n",
            "Epoch 67/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6161.5874 - mae: 6161.5874\n",
            "Epoch 68/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6139.2178 - mae: 6139.2178\n",
            "Epoch 69/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6121.9800 - mae: 6121.9800\n",
            "Epoch 70/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6101.1387 - mae: 6101.1387\n",
            "Epoch 71/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6081.1143 - mae: 6081.1143\n",
            "Epoch 72/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6061.9771 - mae: 6061.9771\n",
            "Epoch 73/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6047.1230 - mae: 6047.1230\n",
            "Epoch 74/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6018.1616 - mae: 6018.1616\n",
            "Epoch 75/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5997.2339 - mae: 5997.2339\n",
            "Epoch 76/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5980.5674 - mae: 5980.5674\n",
            "Epoch 77/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5948.2119 - mae: 5948.2119\n",
            "Epoch 78/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5924.4819 - mae: 5924.4819\n",
            "Epoch 79/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5899.7236 - mae: 5899.7236\n",
            "Epoch 80/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5871.2173 - mae: 5871.2173\n",
            "Epoch 81/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5844.7861 - mae: 5844.7861\n",
            "Epoch 82/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5815.6357 - mae: 5815.6357\n",
            "Epoch 83/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5786.6313 - mae: 5786.6313\n",
            "Epoch 84/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5752.6270 - mae: 5752.6270\n",
            "Epoch 85/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5725.9873 - mae: 5725.9873\n",
            "Epoch 86/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5690.6416 - mae: 5690.6416\n",
            "Epoch 87/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5651.7383 - mae: 5651.7383\n",
            "Epoch 88/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5615.7476 - mae: 5615.7476\n",
            "Epoch 89/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5576.2285 - mae: 5576.2285\n",
            "Epoch 90/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5533.3398 - mae: 5533.3398\n",
            "Epoch 91/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5495.0825 - mae: 5495.0825\n",
            "Epoch 92/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5445.1992 - mae: 5445.1992\n",
            "Epoch 93/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5401.7173 - mae: 5401.7173\n",
            "Epoch 94/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5352.8521 - mae: 5352.8521\n",
            "Epoch 95/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5300.8760 - mae: 5300.8760\n",
            "Epoch 96/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5252.5835 - mae: 5252.5835\n",
            "Epoch 97/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5197.7744 - mae: 5197.7744\n",
            "Epoch 98/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5133.1685 - mae: 5133.1685\n",
            "Epoch 99/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5073.5098 - mae: 5073.5098\n",
            "Epoch 100/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5015.2588 - mae: 5015.2588\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba27a7cd930>"
            ]
          },
          "metadata": {},
          "execution_count": 79
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate the larger model\n",
        "insurance_model_2.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GO7y6Ays8e4B",
        "outputId": "e1fd5ecc-be58-496e-fb2f-42b28adc75d1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 4882.7759 - mae: 4882.7759\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[4882.77587890625, 4882.77587890625]"
            ]
          },
          "metadata": {},
          "execution_count": 80
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "insurance_model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I322ttJu8rKS",
        "outputId": "8272836b-28b9-45f4-a58a-4e1501e83e5e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 7668.9380 - mae: 7668.9380\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[7668.93798828125, 7668.93798828125]"
            ]
          },
          "metadata": {},
          "execution_count": 81
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Set random seed\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create the model (same as above)\n",
        "insurance_model_3 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(100),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "insurance_model_3.compile(loss=tf.keras.losses.mae,\n",
        "                          optimizer=tf.keras.optimizers.Adam(),\n",
        "                          metrics=[\"mae\"])\n",
        "\n",
        "# 3. Fit the model\n",
        "history = insurance_model_3.fit(X_train, y_train, epochs=200)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qSch_-wh9bfZ",
        "outputId": "dcad27e7-bdc4-420a-c5ea-c5f5d1f5ca2d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/200\n",
            "34/34 [==============================] - 1s 2ms/step - loss: 13266.0449 - mae: 13266.0449\n",
            "Epoch 2/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13089.2373 - mae: 13089.2373\n",
            "Epoch 3/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12713.0156 - mae: 12713.0156\n",
            "Epoch 4/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 11985.7598 - mae: 11985.7598\n",
            "Epoch 5/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 10803.7900 - mae: 10803.7900\n",
            "Epoch 6/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 9350.8750 - mae: 9350.8750\n",
            "Epoch 7/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8075.5215 - mae: 8075.5215\n",
            "Epoch 8/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7514.5332 - mae: 7514.5332\n",
            "Epoch 9/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7428.5518 - mae: 7428.5518\n",
            "Epoch 10/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7409.1787 - mae: 7409.1787\n",
            "Epoch 11/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7390.8740 - mae: 7390.8740\n",
            "Epoch 12/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7368.8164 - mae: 7368.8164\n",
            "Epoch 13/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7348.3999 - mae: 7348.3999\n",
            "Epoch 14/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7326.3193 - mae: 7326.3193\n",
            "Epoch 15/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7307.7637 - mae: 7307.7637\n",
            "Epoch 16/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7285.4634 - mae: 7285.4634\n",
            "Epoch 17/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7265.1226 - mae: 7265.1226\n",
            "Epoch 18/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7242.2891 - mae: 7242.2891\n",
            "Epoch 19/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7220.4111 - mae: 7220.4111\n",
            "Epoch 20/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7197.2085 - mae: 7197.2085\n",
            "Epoch 21/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7178.7266 - mae: 7178.7266\n",
            "Epoch 22/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7151.2705 - mae: 7151.2705\n",
            "Epoch 23/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7126.6475 - mae: 7126.6475\n",
            "Epoch 24/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7101.9971 - mae: 7101.9971\n",
            "Epoch 25/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7084.7329 - mae: 7084.7329\n",
            "Epoch 26/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7052.6548 - mae: 7052.6548\n",
            "Epoch 27/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7024.9702 - mae: 7024.9702\n",
            "Epoch 28/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6997.5522 - mae: 6997.5522\n",
            "Epoch 29/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6969.8481 - mae: 6969.8481\n",
            "Epoch 30/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6943.7642 - mae: 6943.7642\n",
            "Epoch 31/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6913.1406 - mae: 6913.1406\n",
            "Epoch 32/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6886.5898 - mae: 6886.5898\n",
            "Epoch 33/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6856.2397 - mae: 6856.2397\n",
            "Epoch 34/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6824.8027 - mae: 6824.8027\n",
            "Epoch 35/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6792.9404 - mae: 6792.9404\n",
            "Epoch 36/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6759.7705 - mae: 6759.7705\n",
            "Epoch 37/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6724.4829 - mae: 6724.4829\n",
            "Epoch 38/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6694.2461 - mae: 6694.2461\n",
            "Epoch 39/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6657.2349 - mae: 6657.2349\n",
            "Epoch 40/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6623.3662 - mae: 6623.3662\n",
            "Epoch 41/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6590.6978 - mae: 6590.6978\n",
            "Epoch 42/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6564.5571 - mae: 6564.5571\n",
            "Epoch 43/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6534.6821 - mae: 6534.6821\n",
            "Epoch 44/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6510.3433 - mae: 6510.3433\n",
            "Epoch 45/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6494.2070 - mae: 6494.2070\n",
            "Epoch 46/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6477.3867 - mae: 6477.3867\n",
            "Epoch 47/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6460.4399 - mae: 6460.4399\n",
            "Epoch 48/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6445.4868 - mae: 6445.4868\n",
            "Epoch 49/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6431.9209 - mae: 6431.9209\n",
            "Epoch 50/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6417.8633 - mae: 6417.8633\n",
            "Epoch 51/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6404.4014 - mae: 6404.4014\n",
            "Epoch 52/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6393.3779 - mae: 6393.3779\n",
            "Epoch 53/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6378.8052 - mae: 6378.8052\n",
            "Epoch 54/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6365.0381 - mae: 6365.0381\n",
            "Epoch 55/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6351.5737 - mae: 6351.5737\n",
            "Epoch 56/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6337.4482 - mae: 6337.4482\n",
            "Epoch 57/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6324.6729 - mae: 6324.6729\n",
            "Epoch 58/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6310.5005 - mae: 6310.5005\n",
            "Epoch 59/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6295.5283 - mae: 6295.5283\n",
            "Epoch 60/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6284.8472 - mae: 6284.8472\n",
            "Epoch 61/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6265.6187 - mae: 6265.6187\n",
            "Epoch 62/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6252.9292 - mae: 6252.9292\n",
            "Epoch 63/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 6234.5264 - mae: 6234.5264\n",
            "Epoch 64/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6217.7002 - mae: 6217.7002\n",
            "Epoch 65/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6200.8735 - mae: 6200.8735\n",
            "Epoch 66/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6183.6006 - mae: 6183.6006\n",
            "Epoch 67/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6170.8369 - mae: 6170.8369\n",
            "Epoch 68/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6148.2441 - mae: 6148.2441\n",
            "Epoch 69/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6132.2642 - mae: 6132.2642\n",
            "Epoch 70/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 6111.9409 - mae: 6111.9409\n",
            "Epoch 71/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 6092.5356 - mae: 6092.5356\n",
            "Epoch 72/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 6074.1484 - mae: 6074.1484\n",
            "Epoch 73/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 6058.1929 - mae: 6058.1929\n",
            "Epoch 74/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6030.8076 - mae: 6030.8076\n",
            "Epoch 75/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6009.4639 - mae: 6009.4639\n",
            "Epoch 76/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5994.1514 - mae: 5994.1514\n",
            "Epoch 77/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5961.9614 - mae: 5961.9614\n",
            "Epoch 78/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5938.8237 - mae: 5938.8237\n",
            "Epoch 79/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5913.8442 - mae: 5913.8442\n",
            "Epoch 80/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5886.5991 - mae: 5886.5991\n",
            "Epoch 81/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5860.7705 - mae: 5860.7705\n",
            "Epoch 82/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5832.4683 - mae: 5832.4683\n",
            "Epoch 83/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5804.1406 - mae: 5804.1406\n",
            "Epoch 84/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 5770.8833 - mae: 5770.8833\n",
            "Epoch 85/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5744.3569 - mae: 5744.3569\n",
            "Epoch 86/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5710.0332 - mae: 5710.0332\n",
            "Epoch 87/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5672.9580 - mae: 5672.9580\n",
            "Epoch 88/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5637.9224 - mae: 5637.9224\n",
            "Epoch 89/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5598.6826 - mae: 5598.6826\n",
            "Epoch 90/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5557.3027 - mae: 5557.3027\n",
            "Epoch 91/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 5520.7437 - mae: 5520.7437\n",
            "Epoch 92/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 5471.2490 - mae: 5471.2490\n",
            "Epoch 93/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 5429.5713 - mae: 5429.5713\n",
            "Epoch 94/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 5383.2651 - mae: 5383.2651\n",
            "Epoch 95/200\n",
            "34/34 [==============================] - 0s 11ms/step - loss: 5330.2778 - mae: 5330.2778\n",
            "Epoch 96/200\n",
            "34/34 [==============================] - 0s 9ms/step - loss: 5284.5283 - mae: 5284.5283\n",
            "Epoch 97/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 5230.3228 - mae: 5230.3228\n",
            "Epoch 98/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 5168.6768 - mae: 5168.6768\n",
            "Epoch 99/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 5109.0322 - mae: 5109.0322\n",
            "Epoch 100/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 5053.6147 - mae: 5053.6147\n",
            "Epoch 101/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 4982.7041 - mae: 4982.7041\n",
            "Epoch 102/200\n",
            "34/34 [==============================] - 0s 7ms/step - loss: 4910.4858 - mae: 4910.4858\n",
            "Epoch 103/200\n",
            "34/34 [==============================] - 0s 7ms/step - loss: 4842.2026 - mae: 4842.2026\n",
            "Epoch 104/200\n",
            "34/34 [==============================] - 0s 9ms/step - loss: 4763.0649 - mae: 4763.0649\n",
            "Epoch 105/200\n",
            "34/34 [==============================] - 0s 8ms/step - loss: 4678.5664 - mae: 4678.5664\n",
            "Epoch 106/200\n",
            "34/34 [==============================] - 0s 7ms/step - loss: 4595.3877 - mae: 4595.3877\n",
            "Epoch 107/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 4509.7837 - mae: 4509.7837\n",
            "Epoch 108/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 4420.4873 - mae: 4420.4873\n",
            "Epoch 109/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 4336.3857 - mae: 4336.3857\n",
            "Epoch 110/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 4251.2246 - mae: 4251.2246\n",
            "Epoch 111/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 4168.4653 - mae: 4168.4653\n",
            "Epoch 112/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 4098.2847 - mae: 4098.2847\n",
            "Epoch 113/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 4030.1753 - mae: 4030.1753\n",
            "Epoch 114/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3986.5247 - mae: 3986.5247\n",
            "Epoch 115/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3944.4055 - mae: 3944.4055\n",
            "Epoch 116/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3919.5886 - mae: 3919.5886\n",
            "Epoch 117/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3895.8403 - mae: 3895.8403\n",
            "Epoch 118/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3872.9998 - mae: 3872.9998\n",
            "Epoch 119/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3854.0386 - mae: 3854.0386\n",
            "Epoch 120/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3836.4597 - mae: 3836.4597\n",
            "Epoch 121/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3829.8604 - mae: 3829.8604\n",
            "Epoch 122/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3824.5432 - mae: 3824.5432\n",
            "Epoch 123/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3819.1487 - mae: 3819.1487\n",
            "Epoch 124/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3809.0889 - mae: 3809.0889\n",
            "Epoch 125/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3797.6174 - mae: 3797.6174\n",
            "Epoch 126/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3809.6096 - mae: 3809.6096\n",
            "Epoch 127/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3799.0249 - mae: 3799.0249\n",
            "Epoch 128/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3793.1545 - mae: 3793.1545\n",
            "Epoch 129/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3802.8376 - mae: 3802.8376\n",
            "Epoch 130/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3787.4534 - mae: 3787.4534\n",
            "Epoch 131/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3783.3777 - mae: 3783.3777\n",
            "Epoch 132/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3776.6572 - mae: 3776.6572\n",
            "Epoch 133/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3772.8108 - mae: 3772.8108\n",
            "Epoch 134/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3772.9407 - mae: 3772.9407\n",
            "Epoch 135/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3769.2483 - mae: 3769.2483\n",
            "Epoch 136/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3767.8057 - mae: 3767.8057\n",
            "Epoch 137/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3777.0706 - mae: 3777.0706\n",
            "Epoch 138/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3786.8516 - mae: 3786.8516\n",
            "Epoch 139/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3763.7004 - mae: 3763.7004\n",
            "Epoch 140/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3764.4460 - mae: 3764.4460\n",
            "Epoch 141/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3764.6104 - mae: 3764.6104\n",
            "Epoch 142/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3764.1118 - mae: 3764.1118\n",
            "Epoch 143/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3756.4485 - mae: 3756.4485\n",
            "Epoch 144/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3752.0847 - mae: 3752.0847\n",
            "Epoch 145/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3755.0735 - mae: 3755.0735\n",
            "Epoch 146/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3758.1416 - mae: 3758.1416\n",
            "Epoch 147/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3752.9968 - mae: 3752.9968\n",
            "Epoch 148/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3757.1326 - mae: 3757.1326\n",
            "Epoch 149/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3745.2263 - mae: 3745.2263\n",
            "Epoch 150/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3739.8972 - mae: 3739.8972\n",
            "Epoch 151/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3742.3652 - mae: 3742.3652\n",
            "Epoch 152/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3744.0750 - mae: 3744.0750\n",
            "Epoch 153/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3745.2092 - mae: 3745.2092\n",
            "Epoch 154/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3739.3452 - mae: 3739.3452\n",
            "Epoch 155/200\n",
            "34/34 [==============================] - 0s 4ms/step - loss: 3740.8655 - mae: 3740.8655\n",
            "Epoch 156/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3736.8276 - mae: 3736.8276\n",
            "Epoch 157/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3735.5747 - mae: 3735.5747\n",
            "Epoch 158/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3730.6724 - mae: 3730.6724\n",
            "Epoch 159/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3727.6758 - mae: 3727.6758\n",
            "Epoch 160/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3734.7244 - mae: 3734.7244\n",
            "Epoch 161/200\n",
            "34/34 [==============================] - 0s 5ms/step - loss: 3729.9226 - mae: 3729.9226\n",
            "Epoch 162/200\n",
            "34/34 [==============================] - 0s 8ms/step - loss: 3726.2339 - mae: 3726.2339\n",
            "Epoch 163/200\n",
            "34/34 [==============================] - 0s 9ms/step - loss: 3724.6733 - mae: 3724.6733\n",
            "Epoch 164/200\n",
            "34/34 [==============================] - 0s 6ms/step - loss: 3727.8689 - mae: 3727.8689\n",
            "Epoch 165/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3717.5818 - mae: 3717.5818\n",
            "Epoch 166/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3721.5735 - mae: 3721.5735\n",
            "Epoch 167/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3722.1682 - mae: 3722.1682\n",
            "Epoch 168/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3716.7908 - mae: 3716.7908\n",
            "Epoch 169/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3715.1104 - mae: 3715.1104\n",
            "Epoch 170/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3709.5635 - mae: 3709.5635\n",
            "Epoch 171/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3709.3645 - mae: 3709.3645\n",
            "Epoch 172/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3710.8530 - mae: 3710.8530\n",
            "Epoch 173/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3706.9875 - mae: 3706.9875\n",
            "Epoch 174/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3711.2791 - mae: 3711.2791\n",
            "Epoch 175/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3710.3401 - mae: 3710.3401\n",
            "Epoch 176/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3708.6392 - mae: 3708.6392\n",
            "Epoch 177/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3701.5344 - mae: 3701.5344\n",
            "Epoch 178/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3701.7629 - mae: 3701.7629\n",
            "Epoch 179/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3707.6123 - mae: 3707.6123\n",
            "Epoch 180/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3710.2234 - mae: 3710.2234\n",
            "Epoch 181/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3697.4919 - mae: 3697.4919\n",
            "Epoch 182/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3695.0273 - mae: 3695.0273\n",
            "Epoch 183/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3693.1172 - mae: 3693.1172\n",
            "Epoch 184/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3699.3479 - mae: 3699.3479\n",
            "Epoch 185/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3694.2847 - mae: 3694.2847\n",
            "Epoch 186/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3696.9736 - mae: 3696.9736\n",
            "Epoch 187/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3690.3374 - mae: 3690.3374\n",
            "Epoch 188/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3694.7402 - mae: 3694.7402\n",
            "Epoch 189/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3684.8308 - mae: 3684.8308\n",
            "Epoch 190/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3684.7922 - mae: 3684.7922\n",
            "Epoch 191/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3699.3542 - mae: 3699.3542\n",
            "Epoch 192/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3686.2107 - mae: 3686.2107\n",
            "Epoch 193/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3676.3892 - mae: 3676.3892\n",
            "Epoch 194/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3677.0657 - mae: 3677.0657\n",
            "Epoch 195/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3674.7429 - mae: 3674.7429\n",
            "Epoch 196/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3682.2644 - mae: 3682.2644\n",
            "Epoch 197/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3667.9602 - mae: 3667.9602\n",
            "Epoch 198/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3672.9592 - mae: 3672.9592\n",
            "Epoch 199/200\n",
            "34/34 [==============================] - 0s 3ms/step - loss: 3681.0205 - mae: 3681.0205\n",
            "Epoch 200/200\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3668.2507 - mae: 3668.2507\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate our third model\n",
        "insurance_model_3.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Y9VOCV4V-VmP",
        "outputId": "bfec80de-0fd3-4f70-b639-5cb03da6c0d9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 3ms/step - loss: 3493.1016 - mae: 3493.1016\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[3493.1015625, 3493.1015625]"
            ]
          },
          "metadata": {},
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "insurance_model.evaluate(X_test, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ShW7DBoK-go-",
        "outputId": "933da917-2b2e-4134-d38a-0809bb2cf7eb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 3ms/step - loss: 7668.9380 - mae: 7668.9380\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[7668.93798828125, 7668.93798828125]"
            ]
          },
          "metadata": {},
          "execution_count": 84
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot history (also known as a loss curve or a training curve)\n",
        "pd.DataFrame(history.history).plot()\n",
        "plt.ylabel(\"loss\")\n",
        "plt.xlabel(\"epochs\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "seth63zG--qS",
        "outputId": "f0f192e6-704a-4ab0-ce3f-61983828ce0d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'epochs')"
            ]
          },
          "metadata": {},
          "execution_count": 85
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        ">**Question:** How long should you train for?\n",
        "\n",
        "It depends. Really... it depends on teh problem you're working for. However, many people have asked this question before... so TensorFlow has a solution! It's called the [EarlyStopping Callback](https://www.tensorflow.org/api_docs/python/tf/keras/callbacks/EarlyStopping), which is a TensorFlow component you can add to your model to stop training once it stops improving a certain metric.\n"
      ],
      "metadata": {
        "id": "a-tuStue_LiS"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Preprocessing data with feature scaling (normalization and standardization)\n",
        "\n",
        "In terms of scaling values, neural networks tend to prefer normalization.\n",
        "\n",
        "If you're not sure on which to sue, you can try both and see which performs better.\n",
        "\n",
        "https://towardsdatascience.com/scale-standardize-or-normalize-with-scikit-learn-6ccc7d176a02"
      ],
      "metadata": {
        "id": "87p0jr3DAbqT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import tensorflow as tf\n",
        "\n",
        "# Read in the insurance dataframe\n",
        "insurance = pd.read_csv(\"https://raw.githubusercontent.com/stedy/Machine-Learning-with-R-datasets/master/insurance.csv\")\n",
        "insurance"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 423
        },
        "id": "LEmyMky_DRCr",
        "outputId": "494b3c06-08c2-4ce1-8f08-edb6d1a64627"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "      age     sex     bmi  children smoker     region      charges\n",
              "0      19  female  27.900         0    yes  southwest  16884.92400\n",
              "1      18    male  33.770         1     no  southeast   1725.55230\n",
              "2      28    male  33.000         3     no  southeast   4449.46200\n",
              "3      33    male  22.705         0     no  northwest  21984.47061\n",
              "4      32    male  28.880         0     no  northwest   3866.85520\n",
              "...   ...     ...     ...       ...    ...        ...          ...\n",
              "1333   50    male  30.970         3     no  northwest  10600.54830\n",
              "1334   18  female  31.920         0     no  northeast   2205.98080\n",
              "1335   18  female  36.850         0     no  southeast   1629.83350\n",
              "1336   21  female  25.800         0     no  southwest   2007.94500\n",
              "1337   61  female  29.070         0    yes  northwest  29141.36030\n",
              "\n",
              "[1338 rows x 7 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ddb99b7e-009f-415b-8308-2d79f7991c06\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>age</th>\n",
              "      <th>sex</th>\n",
              "      <th>bmi</th>\n",
              "      <th>children</th>\n",
              "      <th>smoker</th>\n",
              "      <th>region</th>\n",
              "      <th>charges</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>19</td>\n",
              "      <td>female</td>\n",
              "      <td>27.900</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>southwest</td>\n",
              "      <td>16884.92400</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>18</td>\n",
              "      <td>male</td>\n",
              "      <td>33.770</td>\n",
              "      <td>1</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1725.55230</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>28</td>\n",
              "      <td>male</td>\n",
              "      <td>33.000</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>4449.46200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>33</td>\n",
              "      <td>male</td>\n",
              "      <td>22.705</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>21984.47061</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>32</td>\n",
              "      <td>male</td>\n",
              "      <td>28.880</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>3866.85520</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1333</th>\n",
              "      <td>50</td>\n",
              "      <td>male</td>\n",
              "      <td>30.970</td>\n",
              "      <td>3</td>\n",
              "      <td>no</td>\n",
              "      <td>northwest</td>\n",
              "      <td>10600.54830</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1334</th>\n",
              "      <td>18</td>\n",
              "      <td>female</td>\n",
              "      <td>31.920</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>northeast</td>\n",
              "      <td>2205.98080</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1335</th>\n",
              "      <td>18</td>\n",
              "      <td>female</td>\n",
              "      <td>36.850</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>southeast</td>\n",
              "      <td>1629.83350</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1336</th>\n",
              "      <td>21</td>\n",
              "      <td>female</td>\n",
              "      <td>25.800</td>\n",
              "      <td>0</td>\n",
              "      <td>no</td>\n",
              "      <td>southwest</td>\n",
              "      <td>2007.94500</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1337</th>\n",
              "      <td>61</td>\n",
              "      <td>female</td>\n",
              "      <td>29.070</td>\n",
              "      <td>0</td>\n",
              "      <td>yes</td>\n",
              "      <td>northwest</td>\n",
              "      <td>29141.36030</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>1338 rows × 7 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ddb99b7e-009f-415b-8308-2d79f7991c06')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-ddb99b7e-009f-415b-8308-2d79f7991c06 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-ddb99b7e-009f-415b-8308-2d79f7991c06');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-07e635e7-9713-488f-8570-81fcff40f949\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-07e635e7-9713-488f-8570-81fcff40f949')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-07e635e7-9713-488f-8570-81fcff40f949 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_dc4a42f9-7b72-4a2b-a1d7-184edbbeeed1\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('insurance')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_dc4a42f9-7b72-4a2b-a1d7-184edbbeeed1 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('insurance');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "insurance",
              "summary": "{\n  \"name\": \"insurance\",\n  \"rows\": 1338,\n  \"fields\": [\n    {\n      \"column\": \"age\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 14,\n        \"min\": 18,\n        \"max\": 64,\n        \"num_unique_values\": 47,\n        \"samples\": [\n          21,\n          45,\n          36\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sex\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"male\",\n          \"female\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"bmi\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 6.098186911679014,\n        \"min\": 15.96,\n        \"max\": 53.13,\n        \"num_unique_values\": 548,\n        \"samples\": [\n          23.18,\n          26.885\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"children\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 0,\n        \"max\": 5,\n        \"num_unique_values\": 6,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"smoker\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"no\",\n          \"yes\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"region\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"southeast\",\n          \"northeast\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"charges\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 12110.011236694001,\n        \"min\": 1121.8739,\n        \"max\": 63770.42801,\n        \"num_unique_values\": 1337,\n        \"samples\": [\n          8688.85885,\n          5708.867\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "To prepare our data, we can borrow a few classes from Scikit-learn."
      ],
      "metadata": {
        "id": "k8c8ewkUCJRr"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.compose import make_column_transformer\n",
        "from sklearn.preprocessing import MinMaxScaler, OneHotEncoder\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "# Create a column transformer\n",
        "ct = make_column_transformer(\n",
        "    (MinMaxScaler(), [\"age\", \"bmi\", \"children\"]), # turn all values in these columns between 0 and 1\n",
        "    (OneHotEncoder(handle_unknown=\"ignore\"), [\"sex\", \"smoker\", \"region\"]) # ignore columns it doesn't know about\n",
        ")\n",
        "\n",
        "# Create X & y\n",
        "X = insurance.drop(\"charges\", axis=1)\n",
        "y = insurance[\"charges\"]\n",
        "\n",
        "# Build our train and test sets\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# Fit the column transformer to our training data\n",
        "ct.fit(X_train)\n",
        "\n",
        "# Transform training and test data with normalization (MinMaxScalar) and OneHotEncoder\n",
        "X_train_normal = ct.transform(X_train)\n",
        "X_test_normal = ct.transform(X_test)"
      ],
      "metadata": {
        "id": "K6KIf2ysBC-K"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# What does our data look like now?\n",
        "X_train.loc[0], X_train_normal[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YnMEci9uF1v3",
        "outputId": "43101c6d-7b25-48a9-e5fc-3f8dbe3ded39"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(age                19\n",
              " sex            female\n",
              " bmi              27.9\n",
              " children            0\n",
              " smoker            yes\n",
              " region      southwest\n",
              " Name: 0, dtype: object,\n",
              " array([0.60869565, 0.10734463, 0.4       , 1.        , 0.        ,\n",
              "        1.        , 0.        , 0.        , 1.        , 0.        ,\n",
              "        0.        ]))"
            ]
          },
          "metadata": {},
          "execution_count": 88
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Normalized and one hot encoded data has added columns\n",
        "X_train.shape, X_train_normal.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wuz0ZIlUF7yh",
        "outputId": "c55a783b-95bf-449d-eb61-1210fd9cd84e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((1070, 6), (1070, 11))"
            ]
          },
          "metadata": {},
          "execution_count": 89
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Beautiful! Our data has been normalized and one hot encoded. Now let's biuld a neural network model on it and see how it goes..."
      ],
      "metadata": {
        "id": "YKfo3RaRGbYw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Build a neural network model to fit on our noramlized data\n",
        "tf.random.set_seed(42)\n",
        "\n",
        "# 1. Create the model\n",
        "insurance_model_4 = tf.keras.Sequential([\n",
        "    tf.keras.layers.Dense(100),\n",
        "    tf.keras.layers.Dense(10),\n",
        "    tf.keras.layers.Dense(1)\n",
        "])\n",
        "\n",
        "# 2. Compile the model\n",
        "insurance_model_4.compile(loss=tf.keras.losses.mae,\n",
        "                          optimizer=tf.keras.optimizers.Adam(),\n",
        "                          metrics=[\"mae\"])\n",
        "\n",
        "# Fit the model\n",
        "insurance_model_4.fit(X_train_normal, y_train, epochs=100)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kr61gB7xGjnx",
        "outputId": "8122c286-76c0-467b-8274-ef540e368079"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/100\n",
            "34/34 [==============================] - 1s 2ms/step - loss: 13344.1787 - mae: 13344.1787\n",
            "Epoch 2/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13336.7139 - mae: 13336.7139\n",
            "Epoch 3/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13317.5049 - mae: 13317.5049\n",
            "Epoch 4/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13276.4600 - mae: 13276.4600\n",
            "Epoch 5/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13202.4941 - mae: 13202.4941\n",
            "Epoch 6/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 13084.5449 - mae: 13084.5449\n",
            "Epoch 7/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12912.2314 - mae: 12912.2314\n",
            "Epoch 8/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12675.3008 - mae: 12675.3008\n",
            "Epoch 9/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 12363.4131 - mae: 12363.4131\n",
            "Epoch 10/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 11970.7666 - mae: 11970.7666\n",
            "Epoch 11/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 11504.6426 - mae: 11504.6426\n",
            "Epoch 12/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 10999.2461 - mae: 10999.2461\n",
            "Epoch 13/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 10497.7314 - mae: 10497.7314\n",
            "Epoch 14/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 9999.5264 - mae: 9999.5264\n",
            "Epoch 15/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 9528.1016 - mae: 9528.1016\n",
            "Epoch 16/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 9107.1758 - mae: 9107.1758\n",
            "Epoch 17/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8756.0264 - mae: 8756.0264\n",
            "Epoch 18/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8472.5107 - mae: 8472.5107\n",
            "Epoch 19/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8252.4346 - mae: 8252.4346\n",
            "Epoch 20/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 8102.8179 - mae: 8102.8179\n",
            "Epoch 21/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7991.4673 - mae: 7991.4673\n",
            "Epoch 22/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7915.7671 - mae: 7915.7671\n",
            "Epoch 23/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7856.5981 - mae: 7856.5981\n",
            "Epoch 24/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7803.8755 - mae: 7803.8755\n",
            "Epoch 25/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7765.3369 - mae: 7765.3369\n",
            "Epoch 26/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7714.0239 - mae: 7714.0239\n",
            "Epoch 27/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7672.2544 - mae: 7672.2544\n",
            "Epoch 28/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7630.2554 - mae: 7630.2554\n",
            "Epoch 29/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7588.3608 - mae: 7588.3608\n",
            "Epoch 30/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7545.3916 - mae: 7545.3916\n",
            "Epoch 31/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7502.0181 - mae: 7502.0181\n",
            "Epoch 32/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7458.5303 - mae: 7458.5303\n",
            "Epoch 33/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7414.9077 - mae: 7414.9077\n",
            "Epoch 34/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7367.3652 - mae: 7367.3652\n",
            "Epoch 35/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7321.3262 - mae: 7321.3262\n",
            "Epoch 36/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7272.0190 - mae: 7272.0190\n",
            "Epoch 37/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7222.5059 - mae: 7222.5059\n",
            "Epoch 38/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7172.2456 - mae: 7172.2456\n",
            "Epoch 39/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7118.4775 - mae: 7118.4775\n",
            "Epoch 40/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7064.1016 - mae: 7064.1016\n",
            "Epoch 41/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 7008.6826 - mae: 7008.6826\n",
            "Epoch 42/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6950.8799 - mae: 6950.8799\n",
            "Epoch 43/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6889.5811 - mae: 6889.5811\n",
            "Epoch 44/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6824.6611 - mae: 6824.6611\n",
            "Epoch 45/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6758.9644 - mae: 6758.9644\n",
            "Epoch 46/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6691.1357 - mae: 6691.1357\n",
            "Epoch 47/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6621.7666 - mae: 6621.7666\n",
            "Epoch 48/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6544.8105 - mae: 6544.8105\n",
            "Epoch 49/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6467.8027 - mae: 6467.8027\n",
            "Epoch 50/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6384.2939 - mae: 6384.2939\n",
            "Epoch 51/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6302.0103 - mae: 6302.0103\n",
            "Epoch 52/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6210.9697 - mae: 6210.9697\n",
            "Epoch 53/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6117.3242 - mae: 6117.3242\n",
            "Epoch 54/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 6021.1152 - mae: 6021.1152\n",
            "Epoch 55/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5916.5273 - mae: 5916.5273\n",
            "Epoch 56/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5811.0693 - mae: 5811.0693\n",
            "Epoch 57/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5698.0928 - mae: 5698.0928\n",
            "Epoch 58/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5581.4238 - mae: 5581.4238\n",
            "Epoch 59/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5465.9004 - mae: 5465.9004\n",
            "Epoch 60/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5345.8340 - mae: 5345.8340\n",
            "Epoch 61/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5220.0806 - mae: 5220.0806\n",
            "Epoch 62/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 5093.7192 - mae: 5093.7192\n",
            "Epoch 63/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4966.3110 - mae: 4966.3110\n",
            "Epoch 64/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4835.9341 - mae: 4835.9341\n",
            "Epoch 65/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4708.1387 - mae: 4708.1387\n",
            "Epoch 66/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4578.4458 - mae: 4578.4458\n",
            "Epoch 67/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4464.5894 - mae: 4464.5894\n",
            "Epoch 68/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4355.3252 - mae: 4355.3252\n",
            "Epoch 69/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4251.3906 - mae: 4251.3906\n",
            "Epoch 70/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4154.7476 - mae: 4154.7476\n",
            "Epoch 71/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 4062.8206 - mae: 4062.8206\n",
            "Epoch 72/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3981.5654 - mae: 3981.5654\n",
            "Epoch 73/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3909.7058 - mae: 3909.7058\n",
            "Epoch 74/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3850.3838 - mae: 3850.3838\n",
            "Epoch 75/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3803.4211 - mae: 3803.4211\n",
            "Epoch 76/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3765.8848 - mae: 3765.8848\n",
            "Epoch 77/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3736.5112 - mae: 3736.5112\n",
            "Epoch 78/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3715.8560 - mae: 3715.8560\n",
            "Epoch 79/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3700.1035 - mae: 3700.1035\n",
            "Epoch 80/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3689.1389 - mae: 3689.1389\n",
            "Epoch 81/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3681.1362 - mae: 3681.1362\n",
            "Epoch 82/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3676.7668 - mae: 3676.7668\n",
            "Epoch 83/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3669.6548 - mae: 3669.6548\n",
            "Epoch 84/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3666.0420 - mae: 3666.0420\n",
            "Epoch 85/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3663.1711 - mae: 3663.1711\n",
            "Epoch 86/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3661.4976 - mae: 3661.4976\n",
            "Epoch 87/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3658.1399 - mae: 3658.1399\n",
            "Epoch 88/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3655.6121 - mae: 3655.6121\n",
            "Epoch 89/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3654.0518 - mae: 3654.0518\n",
            "Epoch 90/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3652.1196 - mae: 3652.1196\n",
            "Epoch 91/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3649.2959 - mae: 3649.2959\n",
            "Epoch 92/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3648.4941 - mae: 3648.4941\n",
            "Epoch 93/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3646.4558 - mae: 3646.4558\n",
            "Epoch 94/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3644.7188 - mae: 3644.7188\n",
            "Epoch 95/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3645.6045 - mae: 3645.6045\n",
            "Epoch 96/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3642.4629 - mae: 3642.4629\n",
            "Epoch 97/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3639.9658 - mae: 3639.9658\n",
            "Epoch 98/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3637.9048 - mae: 3637.9048\n",
            "Epoch 99/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3637.5334 - mae: 3637.5334\n",
            "Epoch 100/100\n",
            "34/34 [==============================] - 0s 2ms/step - loss: 3635.9204 - mae: 3635.9204\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x7ba27ac68670>"
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Evaluate our insurance model trained on normalized data\n",
        "insurance_model_4.evaluate(X_test_normal, y_test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "S_w-ox6kJPIQ",
        "outputId": "bff3e679-8fe1-4d13-d3d8-4048824a4c12"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "9/9 [==============================] - 0s 2ms/step - loss: 3439.4670 - mae: 3439.4670\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[3439.467041015625, 3439.467041015625]"
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Insurance model 2 results\n",
        "# 9/9 [==============================] - 0s 4ms/step - loss: 4732.4512 - mae: 4732.4512"
      ],
      "metadata": {
        "id": "kc_gAp_IJffM"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}