Image Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset Processing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n",
      "c:\\users\\chesh\\appdata\\local\\programs\\python\\python36\\lib\\importlib\\_bootstrap.py:219: RuntimeWarning: numpy.ufunc size changed, may indicate binary incompatibility. Expected 192 from C header, got 216 from PyObject\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "from keras.preprocessing import image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "p = Path(\"Images/\")\n",
    "dirs = p.glob(\"*\")\n",
    "labels_dict = {'cat':0, 'dog':1, 'horse':2, 'human': 3}\n",
    "\n",
    "image_data = []\n",
    "labels = []    \n",
    "\n",
    "for folder_dir in dirs:\n",
    "    label = str(folder_dir).split(\"\\\\\")[-1][:-1]\n",
    "    \n",
    "    for img_path in folder_dir.glob(\"*.jpg\"):\n",
    "        img = image.load_img(img_path, target_size=(32,32))\n",
    "        img_array = image.img_to_array(img)\n",
    "        image_data.append(img_array)\n",
    "        labels.append(labels_dict[label])\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "808\n",
      "808\n"
     ]
    }
   ],
   "source": [
    "print(len(image_data))\n",
    "print(len(labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(808, 32, 32, 3) (808,)\n"
     ]
    }
   ],
   "source": [
    "## Convert data into numpy array\n",
    "\n",
    "image_data = np.array(image_data, dtype='float32')/255.0\n",
    "labels = np.array(labels)\n",
    "\n",
    "print(image_data.shape, labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Randomly shuffle data\n",
    "\n",
    "import random \n",
    "combined = list(zip(image_data, labels))\n",
    "random.shuffle(combined)\n",
    "\n",
    "image_data[:], labels[:] = zip(*combined)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Visualize the data\n",
    "\n",
    "def drawImg(img):\n",
    "    import matplotlib.pyplot as plt\n",
    "    plt.imshow(img)\n",
    "    plt.axis('off')\n",
    "    plt.show()\n",
    "    return\n",
    "\n",
    "for i in range(10):\n",
    "    drawImg(image_data[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVM Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class SVM:\n",
    "    def __init__(self,C=1.0):\n",
    "        self.C = C\n",
    "        self.W = 0\n",
    "        self.b = 0\n",
    "        \n",
    "    def hingeLoss(self,W,b,X,Y):\n",
    "        loss  = 0.0\n",
    "        \n",
    "        loss += .5*np.dot(W,W.T)\n",
    "        \n",
    "        m = X.shape[0]\n",
    "        \n",
    "        for i in range(m):\n",
    "            ti = Y[i]*(np.dot(W,X[i].T)+b)\n",
    "            loss += self.C *max(0,(1-ti))\n",
    "            \n",
    "        return loss[0][0]\n",
    "    \n",
    "    def fit(self,X,Y,batch_size=50,learning_rate=0.001,maxItr=500):\n",
    "        \n",
    "        no_of_features = X.shape[1]\n",
    "        no_of_samples = X.shape[0]\n",
    "        \n",
    "        n = learning_rate\n",
    "        c = self.C\n",
    "        \n",
    "        #Init the model parameters\n",
    "        W = np.zeros((1,no_of_features))\n",
    "        bias = 0\n",
    "        \n",
    "        #Initial Loss\n",
    "        \n",
    "        #Training from here...\n",
    "        # Weight and Bias update rule\n",
    "        losses = []\n",
    "        \n",
    "        for i in range(maxItr):\n",
    "            #Training Loop\n",
    "            \n",
    "            l = self.hingeLoss(W,bias,X,Y)\n",
    "            losses.append(l)\n",
    "            ids = np.arange(no_of_samples)\n",
    "            np.random.shuffle(ids)\n",
    "            \n",
    "            #Batch Gradient Descent(Paper) with random shuffling\n",
    "            for batch_start in range(0,no_of_samples,batch_size):\n",
    "                #Assume 0 gradient for the batch\n",
    "                gradw = 0\n",
    "                gradb = 0\n",
    "                \n",
    "                #Iterate over all examples in the mini batch\n",
    "                for j in range(batch_start,batch_start+batch_size):\n",
    "                    if j<no_of_samples:\n",
    "                        i = ids[j]\n",
    "                        ti =  Y[i]*(np.dot(W,X[i].T)+bias)\n",
    "                        \n",
    "                        if ti>1:\n",
    "                            gradw += 0\n",
    "                            gradb += 0\n",
    "                        else:\n",
    "                            gradw += c*Y[i]*X[i]\n",
    "                            gradb += c*Y[i]\n",
    "                            \n",
    "                #Gradient for the batch is ready! Update W,B\n",
    "                W = W - n*W + n*gradw\n",
    "                bias = bias + n*gradb\n",
    "                \n",
    "        \n",
    "        self.W = W\n",
    "        self.b = bias\n",
    "        return W,bias,losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(808, 3072)\n",
      "(808,)\n"
     ]
    }
   ],
   "source": [
    "## Data conversion for One vs One classification\n",
    "\n",
    "M = image_data.shape[0]\n",
    "image_data = image_data.reshape(M,-1)\n",
    "print(image_data.shape)\n",
    "print(labels.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "number_of_classes = len(np.unique(labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def classWiseData(x, y):\n",
    "    data = {}\n",
    "    \n",
    "    for i in range(number_of_classes):\n",
    "        data[i] = []\n",
    "        \n",
    "    for i in range(x.shape[0]):\n",
    "        data[y[i]].append(x[i])\n",
    "        \n",
    "    for k in data.keys():\n",
    "        data[k] = np.array(data[k])\n",
    "        \n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = classWiseData(image_data, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "202\n",
      "202\n",
      "202\n",
      "202\n"
     ]
    }
   ],
   "source": [
    "print(data[0].shape[0])\n",
    "print(data[1].shape[0])\n",
    "print(data[2].shape[0])\n",
    "print(data[3].shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"Combines data of two classes into a single matrix\"\"\"\n",
    "def getDataPairForSVM(d1,d2):\n",
    "    \n",
    "    l1,l2 = d1.shape[0], d2.shape[0]\n",
    "    samples = l1+l2\n",
    "    features = d1.shape[1]\n",
    "    \n",
    "    data_pair = np.zeros((samples,features))\n",
    "    data_labels = np.zeros((samples,))\n",
    "    \n",
    "    data_pair[:l1,:] = d1\n",
    "    data_pair[l1:,:] = d2\n",
    "    \n",
    "    data_labels[:l1] = -1\n",
    "    data_labels[l1:] = 1\n",
    "    \n",
    "    return data_pair, data_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training nC2 SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2216894f1d0>]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mySVM = SVM()\n",
    "xp, yp = getDataPairForSVM(data[0], data[1])\n",
    "w,b,loss = mySVM.fit(xp,yp,learning_rate=0.00001,maxItr=1000)\n",
    "plt.plot(loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def trainSVMs(x,y):\n",
    "    svm_classifiers = {}\n",
    "    \n",
    "    for i in range(number_of_classes):\n",
    "        svm_classifiers[i] = {}\n",
    "        for j in range(i+1, number_of_classes):\n",
    "            xpair,ypair = getDataPairForSVM(data[i],data[j])\n",
    "            wts,b,loss = mySVM.fit(xpair, ypair,learning_rate=0.00001,maxItr=1000)\n",
    "            svm_classifiers[i][j] = (wts,b)\n",
    "            \n",
    "            plt.plot(loss)\n",
    "            plt.show()\n",
    "            \n",
    "    return svm_classifiers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "svm_classifiers = trainSVMs(image_data, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 3072)\n",
      "0.011060000000000025\n"
     ]
    }
   ],
   "source": [
    "cats_dogs = svm_classifiers[0][1]\n",
    "cats_humans = svm_classifiers[0][3]\n",
    "print(cats_dogs[0].shape)\n",
    "print(cats_dogs[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def binaryPredict(x,w,b):\n",
    "    z = np.dot(x,w.T) + b\n",
    "    if z >= 0:\n",
    "        return 1\n",
    "    else:\n",
    "        return -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(x):\n",
    "    \n",
    "    count = np.zeros((number_of_classes,))\n",
    "    \n",
    "    for i in range(number_of_classes):\n",
    "        for j in range(i+1, number_of_classes):\n",
    "            w,b = svm_classifiers[i][j]\n",
    "            \n",
    "            #Take a majority prediction\n",
    "            z = binaryPredict(x,w,b)\n",
    "            \n",
    "            if z==1:\n",
    "                count[j] += 1\n",
    "            else:\n",
    "                count[i] += 1\n",
    "    \n",
    "    final_prediction = np.argmax(count)\n",
    "    return final_prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "0\n"
     ]
    }
   ],
   "source": [
    "print(predict(image_data[0]))\n",
    "print(labels[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(x,y):\n",
    "    \n",
    "    pred = []\n",
    "    count=0\n",
    "    \n",
    "    for i in range(x.shape[0]):\n",
    "        prediction = predict(x[i])\n",
    "        pred.append(prediction)\n",
    "        if prediction==y[i]:\n",
    "            count += 1\n",
    "    \n",
    "    return count/x.shape[0], pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6002475247524752\n"
     ]
    }
   ],
   "source": [
    "acc, ypred = accuracy(image_data, labels)\n",
    "print(acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using scikit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\chesh\\appdata\\local\\programs\\python\\python36\\lib\\importlib\\_bootstrap.py:219: RuntimeWarning: numpy.ufunc size changed, may indicate binary incompatibility. Expected 216, got 192\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "from sklearn import svm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6311881188118812"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "svm_classifier = svm.SVC(kernel='linear', C=1.0)\n",
    "svm_classifier.fit(image_data, labels)\n",
    "ypred_sklearn = svm_classifier.predict(image_data)\n",
    "svm_classifier.score(image_data,labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"Use this method directly - \"\"\"\n",
    "import itertools\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    print(cm)\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SVM classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[111  24  35  32]\n",
      " [ 29 106  34  33]\n",
      " [ 29  22 130  21]\n",
      " [ 25  23  16 138]]\n"
     ]
    }
   ],
   "source": [
    "cnf_matrix = confusion_matrix(labels, ypred)\n",
    "print(cnf_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[111  24  35  32]\n",
      " [ 29 106  34  33]\n",
      " [ 29  22 130  21]\n",
      " [ 25  23  16 138]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cnf_matrix, [0,1,2,3],normalize=False,title='Confusion matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sci-kit Learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[111  30  32  29]\n",
      " [ 22 124  28  28]\n",
      " [ 21  27 134  20]\n",
      " [ 25  21  15 141]]\n"
     ]
    }
   ],
   "source": [
    "cnf_matrix_sklearn = confusion_matrix(labels, ypred_sklearn)\n",
    "print(cnf_matrix_sklearn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n",
      "[[111  30  32  29]\n",
      " [ 22 124  28  28]\n",
      " [ 21  27 134  20]\n",
      " [ 25  21  15 141]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(cnf_matrix_sklearn, [0,1,2,3],normalize=False,title='Confusion matrix',cmap=plt.cm.Blues)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}