Churn Prediction Model

This repository contains a machine learning model designed to predict customer churn in an Iranian telecom company. The project includes model training, evaluation, deployment via Docker, and a RESTful API for serving predictions.

Table of Contents

1. • Installation
2. • Usage
3. • Model Explainability
4. • Docker Instructions
5. • API Instructions
6. • CI/CD Pipeline
7. • License

---

Installation

To set up this project locally, follow these steps:

1. Clone the repository:

    git clone https://github.com/Greenskin44/churn-prediction.git
    cd churn-prediction
   
   
   

2. Usage a. Train and Evaluate the Model To train and evaluate the model, use:

   import joblib
   model = joblib.load('Optimized_IRN_churn_prediction_model.pkl')
   predictions = model.predict(X_test)
   
   

c. Model Explanation Use LIME for explainability on a single instance or SHAP (if supported):

from lime import lime_tabular
explainer = lime_tabular.LimeTabularExplainer(...)
explanation = explainer.explain_instance(X_test.iloc[0],           
model.predict_proba)

3. Model Explainability

   from lime import lime_tabular
   explainer = lime_tabular.LimeTabularExplainer(...)
   explanation = explainer.explain_instance(X_test.iloc[0],model.predict_proba)
   explanation.show_in_notebook()
   
   

4. Docker Deployment a. This project includes Docker setup for consistent deployment. For full Docker instructions, see Docker/README.md.

      docker build -t churn-model .
      docker run -p 5000:5000 churn-model
   
   

5. API Instructions a. The API exposes a /predict endpoint for making predictions. Detailed API usage can be found in API/README.md.

      python app.py
   

c. API Endpoint: POST /predict for prediction with JSON data input.

6. CI/CD Pipeline a. We use GitHub Actions for continuous integration and deployment. For details on the CI/CD setup, see .github/README.md.

7. License a. This project is licensed under the Apache 2.0 License. See the LICENSE file for details.