ImportanceCutoffSelector

class lightautoml.pipelines.selection.importance_based.ImportanceCutoffSelector(feature_pipeline, ml_algo, imp_estimator, fit_on_holdout=True, cutoff=0.0)[source]

Bases: SelectionPipeline

Selector based on importance threshold.

It is important that data which passed to .fit should be ok to fit ml_algo or preprocessing pipeline should be defined.

Parameters:
  • feature_pipeline (Optional[FeaturesPipeline]) – Composition of feature transforms.

  • ml_algo (MLAlgo) – Tuple (MlAlgo, ParamsTuner).

  • imp_estimator (ImportanceEstimator) – Feature importance estimator.

  • fit_on_holdout (bool) – If use the holdout iterator.

  • cutoff (float) – Threshold to cut-off features.

perform_selection(train_valid=None)[source]

Select features based on cutoff value.

Parameters:

train_valid (Optional[TrainValidIterator]) – Not used.