Feature Selection via Dependence Maximization

作     者：Song, Le Smola, Alex Gretton, Arthur Bedo, Justin Borgwardt, Karsten 

作者机构：Georgia Inst Technol Atlanta GA 30332 USA Yahoo Res Santa Clara CA 95053 USA Gatsby Computat Neurosci Unit London WC1N 3AR England Natl ICT Australia Stat Machine Learning Program Canberra ACT 0200 Australia Max Planck Inst Machine Learning & Computat Biol Res Grp D-72076 Tubingen Germany Max Planck Inst Intelligent Syst Grp D-72076 Tubingen Germany Australian Natl Univ Canberra ACT 0200 Australia 

出 版 物：《JOURNAL OF MACHINE LEARNING RESEARCH》 (机器学习研究杂志)

年 卷 期：2012年第13卷第5期

页      面：1393-1434页

核心收录：

学科分类：08[工学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：Australian Government's Baking Australia's Ability initiative  in part through the Australian Research Council 

主　　题：kernel methods feature selection independence measure Hilbert-Schmidt independence criterion Hilbert space embedding of distribution 

摘      要：We introduce a framework for feature selection based on dependence maximization between the selected features and the labels of an estimation problem, using the Hilbert-Schmidt Independence Criterion. The key idea is that good features should be highly dependent on the labels. Our approach leads to a greedy procedure for feature selection. We show that a number of existing feature selectors are special cases of this framework. Experiments on both artificial and real-world data show that our feature selector works well in practice.