FEATURE-SELECTION BASED ON THE APPROXIMATION OF CLASS DENSITIES BY FINITE MIXTURES OF SPECIAL TYPE

被引：44

作者：

PUDIL, P ^{[1
]}

NOVOVICOVA, J ^{[1
]}

CHOAKJARERNWANIT, N ^{[1
]}

KITTLER, J ^{[1
]}

机构：

[1] ACAD SCI CZECH REPUBL,INST INFORMAT THEORY & AUTOMAT,CR-18208 PRAGUE 8,CZECH REPUBLIC

来源：

PATTERN RECOGNITION | 1995年 / 28卷 / 09期

关键词：

FEATURE SELECTION; FEATURE ORDERING; MIXTURE DISTRIBUTION; MAXIMUM LIKELIHOOD; EM ALGORITHM;

D O I：

10.1016/0031-3203(94)00009-B

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

A new method of feature selection based on the approximation of class conditional densities by a mixture of parameterized densities of a special type, suitable especially for multimodal data, is presented. No search procedure is needed when using the proposed method. Its performance is tested both on real and simulated data.

引用

页码：1389 / 1398

页数：10

共 24 条

[1]

DAY NE, 1969, BIOMETRIKA, V56, P463, DOI 10.1093/biomet/56.3.463

[2] MAXIMUM LIKELIHOOD FROM INCOMPLETE DATA VIA EM ALGORITHM [J].

DEMPSTER, AP ;

LAIRD, NM ;

RUBIN, DB .

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-METHODOLOGICAL, 1977, 39 (01) :1-38

[3]

DEVIJVER PA, 1992, PATTERN RECOGNITION

[4] MAXIMUM LIKELIHOOD ESTIMATION FOR MIXTURES OF 2 NORMAL DISTRIBUTIONS [J].

DICK, NP ;

BOWDEN, DC .

BIOMETRICS, 1973, 29 (04) :781-790

[5]

Duda R. O., 1973, PATTERN CLASSIFICATI, V3

[6] A MONTE-CARLO INVESTIGATION OF THE LIKELIHOOD RATIO TEST FOR THE NUMBER OF COMPONENTS IN A MIXTURE OF NORMAL-DISTRIBUTIONS [J].

EVERITT, BS .

MULTIVARIATE BEHAVIORAL RESEARCH, 1981, 16 (02) :171-180

[7] SOME METHODS FOR STUDYING THE MIXTURE OF 2 NORMAL (LOGNORMAL) DISTRIBUTIONS [J].

FOWLKES, EB .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1979, 74 (367) :561-575

[8]

FUKUNAGA K, 1990, INTRO STATISTICAL PA

[9]

GRIM J, 1986, KYBERNETIKA, V22, P142

[10]

GRIM J, 1982, KYBERNETIKA, V18, P173

← 1 2 3 →