Variable selection in model-based discriminant analysis

JOURNAL OF MULTIVARIATE ANALYSIS

卷:102

Article

Maugis, C.¹  Celeux, G.²  Martin-Magniette, M-L^3,4 

[1] Univ Toulouse, INSA Toulouse, Inst Math Toulouse, F-31077 Toulouse 4, France

[2] Inria Saclay Ile de France, Sophia Antipolis, France

[3] UMR AgroParisTech INRA MIA 518, Paris, France

[4] ERL CNRS 8196, UEVE, URGV UMR INRA 1165, Evry, France

关键词: Discriminant, redundant or independent variables;    Variable selection;    Gaussian classification models;    Linear regression;    BIC;   

DOI  :  10.1016/j.jmva.2011.05.004

来源: Elsevier

【 摘 要 】

A general methodology for selecting predictors for Gaussian generative classification models is presented. The problem is regarded as a model selection problem. Three different roles for each possible predictor are considered: a variable can be a relevant classification predictor or not, and the irrelevant classification variables can be linearly dependent on a part of the relevant predictors or independent variables. This variable selection model was inspired by a previous work on variable selection in model-based clustering. A BIC-like model selection criterion is proposed. It is optimized through two embedded forward stepwise variable selection algorithms for classification and linear regression. The model identifiability and the consistency of the variable selection criterion are proved. Numerical experiments on simulated and real data sets illustrate the interest of this variable selection methodology. In particular, it is shown that this well ground variable selection model can be of great interest to improve the classification performance of the quadratic discriminant analysis in a high dimension context. (C) 2011 Elsevier Inc. All rights reserved.

【 授权许可】

Free