aliibtisam1001/customer-churn-prediction — reverse-engineered prompt

Build me a customer churn prediction project for a telecom company that I can use as a data science portfolio piece.

I want it to load the Telco Customer Churn dataset, explore churn patterns with easy charts, clean and prepare the data, handle class imbalance, train a few models, and compare them clearly. Use Logistic Regression, Random Forest, and XGBoost, then show metrics like ROC AUC, F1, precision, recall, and a confusion matrix so I can explain which model is best.

Please also add SHAP explanations so the app can show why a customer is likely to leave, both overall and for one selected customer. Make a simple Streamlit dashboard with multiple pages for overview, data exploration, model results, explanations, and live prediction where someone can enter customer details and get a churn risk.

Keep the code organized, include training scripts, saved models, requirements, and clear run instructions. Look up current docs online if you need to.