Expand description
Recursive Feature Elimination (RFE).
Mirrors sklearn.feature_selection.RFE with a callback-based API: the
caller provides a function that, given (X, y), returns per-feature
importances (e.g. |coef_| for linear models or feature_importances_
for trees). RFE repeatedly drops the step least-important features
until n_features_to_select remain.
Structs§
- Fitted
Rfe - Fitted
Rfecv - Fitted
Sequential Feature Selector - Rfe
- Rfecv
- Recursive Feature Elimination with Cross-Validated selection of the optimal number of features.
- Sequential
Feature Selector