Resampling methods for parameter-free and robust feature selection with mutual information

Damien François (1) , Fabrice Rossi (2) , Vincent Wertz (1) , Michel Verleysen (3)

1 CSAM - Centre for Systems Engineering and Applied Mechanics
2 AxIS - Usage-centered design, analysis and improvement of information systems
3 DICE - MLG - Dispositifs Intégrés et Circuits Electroniques Machine Learning Group

Combining the mutual information criterion with a forward feature selection strategy offers a good trade-off between optimality of the selected feature subset and computation time. However, it requires to set the parameter(s) of the mutual information estimator and to determine when to halt the forward procedure. These two choices are difficult to make because, as the dimensionality of the subset increases, the estimation of the mutual information becomes less and less reliable. This paper proposes to use resampling methods, a K-fold cross-validation and the permutation test, to address both issues. The resampling methods bring information about the variance of the estimator, information which can then be used to automatically set the parameter and to calculate a threshold to stop the forward procedure. The procedure is illustrated on a synthetic dataset as well as on real-world examples.

Mots clés: Mutual information; Permutation test; Feature selection; Resampling

Domaines: Réseau de neurones [cs.NE]; Apprentissage [cs.LG]; Statistiques [math.ST]

Damien François, Fabrice Rossi, Vincent Wertz, Michel Verleysen. Resampling methods for parameter-free and robust feature selection with mutual information. Neurocomputing, 2007, 70 (7-9), pp.1276-1288. ⟨10.1016/j.neucom.2006.11.019⟩. ⟨inria-00174298⟩