Telecom churn model

The brief

Mobile operators lose customers through number portability every month. Predict which customers are likely to leave, compare ensemble models, and explain the reasons for each, so the retention team can make the right offer. Good looks like: careful validation, a threshold chosen by the cost of an offer, and explanations per customer.

What you will build

1. Explore usage, plan and service-call features
2. Train random forest, gradient boosting and stacked models
3. Evaluate with AUC and recall at a budget
4. Explain predictions with SHAP
5. Suggest an offer per customer segment

Free, open data to build it with

• Telecom churn datasets ↗ Kaggle

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.