Expand description
Evaluation-loop utilities that compose a splitter, a model-fitting closure, and one or more scorers.
cross_validate— the day-to-day CV loop, multi-scorer, optionally parallel.nested_cross_validate— outer estimation loop around an inner tuning loop, for a bias-free performance estimate.learning_curve— score vs. training-set size.validation_curve— score vs. a single hyperparameter.
Structs§
- CvResults
- Results of a
cross_validaterun. - Learning
Curve - Output of
learning_curve. - Nested
CvResults - Output of
nested_cross_validate. - Validation
Curve - Output of
validation_curve.
Enums§
- Train
Size - A training-set size, given either absolutely or as a fraction of the largest usable training set.
Functions§
- cross_
validate - Cross-validate a model-fitting closure with one or more scorers.
- learning_
curve - Compute a learning curve: for each training size and each fold, fit on a prefix of the fold’s training set and score on both that prefix and the held-out validation set.
- nested_
cross_ validate - Run nested cross-validation.
- validation_
curve - Compute a validation curve: for each hyperparameter value and each fold, fit with that value and score on train and validation.
Type Aliases§
- Boxed
Scorer - A scorer usable by the evaluation utilities.