Skip to content

Latest commit

 

History

History
109 lines (78 loc) · 2.36 KB

README.md

File metadata and controls

109 lines (78 loc) · 2.36 KB
title emoji colorFrom colorTo sdk app_port tags pinned short_description license
UnsupervisedCustumerPrediction
🧩
indigo
blue
docker
8501
streamlit
false
Streamlit app that predicts cluster labels from uploaded CSV
mit

🧩 Clustering Predictor (KMeans / GMM)

🔗 Live Demo:(https://huggingface.co/spaces/EnYa32/UnsupervisedCustumerPrediction)

📓 Kaggle Competition: https://www.kaggle.com/code/enesyama/unsupervisedcustumer-clustering

⚡ Quick Facts

  • Task: Unsupervised Clustering
  • Models: KMeans + Gaussian Mixture Model
  • Preprocessing: StandardScaler + PCA (95%)
  • Input: Tabular CSV
  • Output: Cluster Labels
  • Deployment: Streamlit App
  • Pipeline: Saved sklearn objects (.pkl)

🧠 Pipeline Structure

Raw Features
→ StandardScaler
→ PCA (95% variance)
→ Clustering Model (KMeans / GMM)
→ Cluster Label Prediction

✅ What this app does

  • Upload a CSV file
  • The app checks required feature columns
  • Applies scaler + PCA
  • Outputs Predicted cluster label for each row
  • Lets you download the predictions as a CSV

📸 Model & Clustering Visualizations

PCA Explained Variance

PCA Variance

KMeans Elbow Method

Elbow Method

Cluster Visualization (PCA 2D)

Clusters

📊 Evaluation

Because this is an unsupervised learning project, evaluation is not based on labeled accuracy.

Model quality was evaluated using:

  • Kaggle leaderboard score
  • Cluster stability
  • PCA visualization
  • Distribution consistency across clusters

🛠️ Tech Stack

  • Python
  • scikit-learn
  • PCA
  • KMeans
  • Gaussian Mixture Models
  • Streamlit
  • Pickle model persistence

📦 Required Model Artifacts

The following trained pipeline artifacts must exist in the repo root:

  • feature_names.pkl
  • scaler.pkl
  • pca.pkl
  • kmeans_model_k9.pkl
  • gmm_model_k9.pkl

🧾 Input format

Your CSV must include all feature columns stored in feature_names.pkl.

Optional:

  • You may include an id or Id column.
    If present, it will be included in the output as Id.

▶️ Run locally

pip install -r requirements.txt
streamlit run app.py


Visual separation in 2D does not always reflect the Kaggle metric.

You can download the CSV file here:
https://www.kaggle.com/competitions/tabular-playground-series-jul-2022/data