Catalogo Articoli (Spogli Riviste)

OPAC HELP

Titolo:
Split and merge EM algorithm for improving Gaussian ixture density estimates
Autore:
Ueda, N; Nakano, R; Ghahramani, Z; Hinton, GE;
Indirizzi:
NTT, Commun Sci Labs, Seika, Kyoto 6190237, Japan NTT Seika Kyoto Japan 6190237 ommun Sci Labs, Seika, Kyoto 6190237, Japan Univ Coll London, Gatsby Computat Neurosci Unit, London WC1N 3AR, England Univ Coll London London England WC1N 3AR Unit, London WC1N 3AR, England
Titolo Testata:
JOURNAL OF VLSI SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY
fascicolo: 1-2, volume: 26, anno: 2000,
pagine: 133 - 140
SICI:
1387-5485(200008)26:1-2<133:SAMEAF>2.0.ZU;2-F
Fonte:
ISI
Lingua:
ENG
Tipo documento:
Article
Natura:
Periodico
Settore Disciplinare:
Engineering, Computing & Technology
Citazioni:
12
Recensione:
Indirizzi per estratti:
Indirizzo: Ueda, N NTT, Commun Sci Labs, Seika, Kyoto 6190237, Japan NTT Seika KyotoJapan 6190237 i Labs, Seika, Kyoto 6190237, Japan
Citazione:
N. Ueda et al., "Split and merge EM algorithm for improving Gaussian ixture density estimates", J VLSI S P, 26(1-2), 2000, pp. 133-140

Abstract

The EM algorithm for Gaussian mixture models often gets caught in local maxima of the likelihood which involve having too many Gaussians in one part of the space and too few in another, widely separated part of the space. Wepresent a new EM algorithm which performs split and merge operations on the Gaussians to escape from these configurations. This algorithm uses two novel criteria for efficiently selecting the split and merge candidates. Experimental results on synthetic and real data show the effectiveness of usingthe split and merge operations to improve the likelihood of both the training data and of held-out test data.

ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 23/01/20 alle ore 04:00:06