Selection of generative models in classification

被引:18
作者
Bouchard, G
Celeux, G
机构
[1] Xerox Res Ctr Europe, F-38240 Meylan, France
[2] Univ Paris 11, Dept Math, F-91405 Orsay, France
关键词
generative classification; integrated likelihood; integrated conditional likelihood; classification entropy; cross-validated error rate; AIC and BIC criteria;
D O I
10.1109/TPAMI.2006.82
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper is concerned with the selection of a generative model for supervised classification. Classical criteria for model selection assess the fit of a model rather than its ability to produce a low classification error rate. A new criterion, the Bayesian Entropy Criterion (BEC), is proposed. This criterion takes into account the decisional purpose of a model by minimizing the integrated classification entropy. It provides an interesting alternative to the cross-validated error rate which is computationally expensive. The asymptotic behavior of the BEC criterion is presented. Numerical experiments on both simulated and real data sets show that BEC performs better than the BIC criterion to select a model minimizing the classification error rate and provides analogous performance to the cross-validated error rate.
引用
收藏
页码:544 / 554
页数:11
相关论文
共 37 条
[1]   NEW LOOK AT STATISTICAL-MODEL IDENTIFICATION [J].
AKAIKE, H .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 1974, AC19 (06) :716-723
[2]  
[Anonymous], LEARNING KERNELS
[3]  
[Anonymous], P 7 EUR C COMP VIS
[4]   Regularized Gaussian discriminant analysis through eigenvalue decomposition [J].
Bensmail, H ;
Celeux, G .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1996, 91 (436) :1743-1748
[5]  
Bernardo J., 2009, Bayesian theory
[6]   Assessing a mixture model for clustering with the integrated completed likelihood [J].
Biernacki, C ;
Celeux, G ;
Govaert, G .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2000, 22 (07) :719-725
[7]  
Blei DM, 2002, ADV NEUR IN, V14, P601
[8]  
BOUCHARD G, 2003, P CLASS DAT AN GROUP, P75
[9]  
BOUCHARD G, 2005, P INT C COMP VIS PAT
[10]  
Csurka G., 2004, WORKSH STAT LEARN CO, V1, P1, DOI DOI 10.1234/12345678