Sparse minimum Redundancy Maximum Relevance for feature selection

TL;DR

This paper introduces a penalized mRMR feature selection method that effectively identifies inactive features while controlling FDR, demonstrated through simulations and real data, offering a more conservative alternative to existing methods.

Contribution

It develops a continuous, penalized version of mRMR with a multi-stage knockoff filter, enabling accurate inactive feature detection and FDR control.

Findings

Performs comparably to HSIC-LASSO in feature selection.

More conservative in the number of features selected.

Requires only an FDR threshold setting.

Abstract

We propose a feature screening method that integrates both feature-feature and feature-target relationships. Inactive features are identified via a penalized minimum Redundancy Maximum Relevance (mRMR) procedure, which is the continuous version of the classic mRMR penalized by a non-convex regularizer, and where the parameters estimated as zero coefficients represent the set of inactive features. We establish the conditions under which zero coefficients are correctly identified to guarantee accurate recovery of inactive features. We introduce a multi-stage procedure based on the knockoff filter enabling the penalized mRMR to discard inactive features while controlling the false discovery rate (FDR). Our method performs comparably to HSIC-LASSO but is more conservative in the number of selected features. It only requires setting an FDR threshold, rather than specifying the number of features to retain. The effectiveness of the method is illustrated through simulations and real-world datasets. The code to reproduce this work is available on the following GitHub: https://github.com/PeterJackNaylor/SmRMR.