AN ENTROPY CRITERION FOR ASSESSING THE NUMBER OF CLUSTERS IN A MIXTURE MODEL

Authors
CELEUX G SOROMENHO G
Citation
G. Celeux et G. Soromenho, AN ENTROPY CRITERION FOR ASSESSING THE NUMBER OF CLUSTERS IN A MIXTURE MODEL, Journal of classification, 13(2), 1996, pp. 195-212
Citations number
33
Categorie Soggetti
Social Sciences, Mathematical Methods","Mathematical, Methods, Social Sciences","Mathematics, Miscellaneous
Journal title
Journal of classification → ACNP
ISSN journal
01764268
Volume
13
Issue
2
Year of publication
1996
Pages
195 - 212
Database
ISI
SICI code
0176-4268(1996)13:2<195:AECFAT>2.0.ZU;2-S
Abstract
In this paper, we consider an entropy criterion to estimate the number of clusters arising from a mixture model. This criterion is derived f rom a relation linking the likelihood and the classification likelihoo d of a mixture. Its performance is investigated through Monte Carlo ex periments, and it shows favorable results compared to other classical criteria.