Which univariate selector score can detect dependencies missed by a linear F-test, and what is the trade-off?
Classic MLLevel: IntermediateNonlinear univariate dependence
Which univariate selector score can detect dependencies missed by a linear F-test, and what is the trade-off?
Hints
- Linear association is not all dependence.
- Flexibility increases estimation demands.
Answer
Mutual-information estimators can detect nonlinear dependence. Their nonparametric estimation generally needs more data and careful handling of discrete versus continuous features.
Source and licence: scikit-learn 1.7.2 — feature selection · BSD-3-Clause · answer: Quantoros editorial