Skip to Main content Skip to Navigation
Journal articles

Stochastic versions of the EM algorithm: An experimental study in the mixture case

Abstract : We compare three different stochastic versions of the EM algorithm: The Stochastic EM algorithm (SEM), the ''Simulated Annealing'' EM algorithm (SAEM) and the Monte Carlo EM algorithm (MCEM). We focus particularly on the mixture of distributions problem. In this context, we investigate the practical behaviour of these algorithms through intensive Monte Carlo numerical simulations and a real data study. We show that, for some particular mixture situations, the SEM algorithm is almost always preferable to the EM and ''simulated annealing'' versions SAEM and MCEM. For some severely overlapping mixtures, however, none of these algorithms can be confidently used. Then, SEM can be used as an efficient data exploratory tool for locating significant maxima of the likelihood function. In the real data case, we show that the SEM stationary distribution provides a contrasted view of the loglikelihood by emphasizing sensible maxima.
Document type :
Journal articles
Complete list of metadatas

https://hal-upec-upem.archives-ouvertes.fr/hal-00693519
Contributor : Admin Lama <>
Submitted on : Wednesday, May 2, 2012 - 5:39:42 PM
Last modification on : Thursday, March 19, 2020 - 12:26:02 PM

Identifiers

Citation

Gilles Celeux, Didier Chauveau, Jean Diebolt. Stochastic versions of the EM algorithm: An experimental study in the mixture case. Journal of Statistical Computation and Simulation, Taylor & Francis, 1996, 55 (4), pp.287--314. ⟨10.1080/00949659608811772⟩. ⟨hal-00693519⟩

Share

Metrics

Record views

544