Catalogo Articoli (Spogli Riviste)

OPAC HELP

Titolo:
EM algorithm with split and merge operations for mixture models
Autore:
Ueda, N; Nakano, R;
Indirizzi:
NTT, Commun Sci Labs, Kyoto 6190237, Japan NTT Kyoto Japan 6190237NTT, Commun Sci Labs, Kyoto 6190237, Japan
Titolo Testata:
IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS
fascicolo: 12, volume: E83D, anno: 2000,
pagine: 2047 - 2055
SICI:
0916-8532(200012)E83D:12<2047:EAWSAM>2.0.ZU;2-Z
Fonte:
ISI
Lingua:
ENG
Keywords:
EM algorithm; split and merge operations; mixture models; maximum likelihood estimates; dimensionality reduction;
Tipo documento:
Article
Natura:
Periodico
Settore Disciplinare:
Engineering, Computing & Technology
Citazioni:
10
Recensione:
Indirizzi per estratti:
Indirizzo: Ueda, N NTT, Commun Sci Labs, Kyoto 6190237, Japan NTT Kyoto Japan 6190237 T, Commun Sci Labs, Kyoto 6190237, Japan
Citazione:
N. Ueda e R. Nakano, "EM algorithm with split and merge operations for mixture models", IEICE T INF, E83D(12), 2000, pp. 2047-2055

Abstract

The maximum likelihood estimate of a mixture model is usually found by using the EM algorithm. However, the EM algorithm suffers from a local optima problem and therefore we cannot obtain the potential performance of mixturemodels in practice. In the case of mixture models, local maxima often havetoo many components of a mixture model in one part of the space and too few in another, widely separated part of the space. To escape from such configurations we proposed a new variant of the Ehl algorithm in which simultaneous split and merge operations are repeatedly performed by using a new criterion for efficiently selecting the split and merge candidates. We apply the proposed algorithm to the training of Gaussian mixtures and the dimensionality reduction based on a mixture of factor analyzers using synthetic and real data and show that the proposed algorithm can markedly improve the ML estimates.

ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 18/01/20 alle ore 02:16:57