Bayesian approaches to Gaussian mixture modeling

被引：206

作者：

Roberts, SJ ^{[1
]}

Husmeier, D ^{[1
]}

Rezek, I ^{[1
]}

Penny, W ^{[1
]}

机构：

[1] Univ London Sch Pharm, Dept Elect & Elect Engn, London, England

来源：

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE | 1998年 / 20卷 / 11期

基金：

英国工程与自然科学研究理事会;

关键词：

cluster analysis; unsupervised learning; Bayesian methods; Gaussian mixture models;

D O I：

10.1109/34.730550

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

A Bayesian-based methodology is presented which automatically penalizes overcomplex models being fitted to unknown data. We show that, with a Gaussian mixture model, the approach is able to select an "optimal" number of components in the model and so partition data sets. The performance of the Bayesian method is compared to other methods of optimal model selection and found to give good results. The methods are tested on synthetic and real data sets.

引用

页码：1133 / 1142

页数：10

共 26 条

[1]

Anderson E., 1935, Bulletin of the American IRIS Society, V59, P2

[2]

BAXTER RA, 1994, 207 MON U DEP COMP S

[3]

Bezdek J. C., 1981, Pattern recognition with fuzzy objective function algorithms

[4]

Bishop C. M., 1995, NEURAL NETWORKS PATT

[5]

Conway J., 2010, Sphere Packings, Lattices and Groups, DOI DOI 10.1007/978-1-4757-2016-7

[6] 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

[7] UNSUPERVISED OPTIMAL FUZZY CLUSTERING [J].

GATH, I ;

GEVA, AB .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1989, 11 (07) :773-781

[8]

HALL P, 1985, J ROY STAT SOC B MET, V47, P155

[9]

HAND DJ, 1994, KERNEL DISCRIMINANT

[10]

Harvey A. C, 1981, ECONOMETRIC ANAL TIM

← 1 2 3 →