medium10 pts · first answer countsdata-leakage
Suspiciously strong churn feature
A churn model reaches 0.99 AUC. The top feature is "cancellation_reason_filled". What is the most likely issue?
Discussion 0
Join free to post in the discussion.
Related questions
- Oversampling before cross-validationData Scientist
- Target encoding without leakageData Scientist
- Great backtest, poor live forecastData Scientist
- Scaling before splittingMachine Learning Engineer
- Target encoding without leakageMachine Learning Engineer
- Several scans per patientMachine Learning Engineer
- A feature that is too goodMachine Learning Engineer
- SMOTE applied before cross-validationMachine Learning Engineer