depynd.feature_selection._mifs()

depynd.feature_selection._mifs(X, y, lamb, k, **kwargs)

Select effective features in X on predicting y using mutual-information-based feature selection [brown2012conditional].

Parameters:
  • X (array-like, shape (n_samples, n_features)) – Observations of feature variables.
  • y (array-like, shape (n_samples)) – Observations of the target variable.
  • lamb (float or None) – Threshold for independence tests. Ignored if k is specified.
  • k (int or None) – Number of selected features.
  • kwargs (dict) – Optional parameters for MI estimation.
Returns:

indices – Indices of the selected features.
Return type:

list

References

[brown2012conditional]Brown, Gavin, et al. “Conditional likelihood maximisation: a unifying framework for information theoretic feature selection.” Journal of machine learning research 13.Jan (2012): 27-66.