This project aims to predict customer churn using machine learning techniques, specifically XGBoost. The goal is to identify customers likely to leave a service, allowing for proactive retention strategies.
The project follows a standard end-to-end machine learning pipeline:
- Data Cleaning – Handled missing values, removed inconsistencies.
- Exploratory Data Analysis (EDA) – Explored distributions, correlations, and key patterns.
- Feature Engineering – Converted categorical variables, normalized numerical features.
- Class Imbalance Handling – Applied techniques like SMOTE to improve recall on minority class.
- Modeling – Trained and tuned an XGBoost classifier for best performance.
- Evaluation – Assessed using accuracy, precision, recall, F1-score, ROC-AUC, and confusion matrix.
- Language: Python 3.x
- Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, XGBoost, imbalanced-learn
| Directory/File | Description |
|---|---|
| data/ | Cleaned dataset |
| notebooks/ | EDA and modeling Jupyter notebooks |
| models/ | XGBoost model |
| README.md | Project overview and instructions |