Skip to Main content Skip to Navigation
Reports

Binary Feature Selection with Conditional Mutual Information

Abstract : In a context of classification, we propose to use conditional mutual information to select a family of binary features which are individually discriminating and weakly dependent. We show that on a task of image classification, despite its simplicity, a naive Bayesian classifier based on features selected with this Conditional Mutual Information Maximization (CMIM) criterion performs as well as a classifier built with AdaBoost. We also show that this classification method is more robust than boosting when trained on a noisy data set.
Document type :
Reports
Complete list of metadata

Cited literature [1 references]  Display  Hide  Download

https://hal.inria.fr/inria-00071638
Contributor : Rapport de Recherche Inria <>
Submitted on : Tuesday, May 23, 2006 - 6:22:42 PM
Last modification on : Friday, May 25, 2018 - 12:02:03 PM
Long-term archiving on: : Sunday, April 4, 2010 - 10:30:26 PM

Identifiers

  • HAL Id : inria-00071638, version 1

Collections

Citation

François Fleuret. Binary Feature Selection with Conditional Mutual Information. [Research Report] RR-4941, INRIA. 2003. ⟨inria-00071638⟩

Share

Metrics

Record views

266

Files downloads

477