Cross-validation and Honest Nesting

Single-split scores wander; stratified K-fold mean/std; SelectKBest before CV leaks; GroupKFold keeps patients whole. Home: five holds spread 0.1; five-fold mean 0.7575; leaky 0.9375 vs honest 0.85; plain fold 0.895 vs whole-patient folds 0.525 (gap 0.37).

This Python for ML lesson is part of a paid plan. Sign in with a plan that includes this course to read it.