XGChurn is a robust machine learning project designed to predict customer churn in the banking sector. Using XGBoost and explainability tools like SHAP, the project achieves high accuracy and provides actionable insights. The project also features an interactive Streamlit dashboard for visualizing predictions and understanding the drivers behind churn.
- Predicts churn with 87% accuracy and 85% ROC-AUC.
- Implements advanced feature engineering (e.g., balance-to-salary ratio).
- Visualizes model explainability using SHAP values.
- Offers a user-friendly Streamlit dashboard for predictions and insights.
- Programming Language: Python
- Machine Learning: XGBoost, scikit-learn
- Dashboard: Streamlit
- Libraries: pandas, numpy, matplotlib, SHAP
- Data Format: CSV (Bank customer data)
- Clone the repository:
git clone https://github.com/your-username/XGChurn.git