Capstone: churn prediction

Why it matters

This is the project the whole phase has been building toward, and the one marketing teams actually ask for: "Who's about to leave, and is it worth trying to keep them?" Doing it well takes everything from Phase 3: clean splits, baselines, pipelines, cross-validation, the right metrics, costs, and an honest account of the model's limits. The result is a plan a marketing director can act on, not just a model.

What you'll be able to do

• Run a complete ML project: framing, baselines, model comparison, one honest test
• Translate model scores into a campaign: lift, a cost-based contact list and its expected value
• Explain what drives the predictions and check whether the probabilities can be trusted
• State what a churn model can't tell you, and how to test a campaign properly

Before this lesson

• The scikit-learn workflow
• Random forests and boosting
• Precision, recall and F1