Stochastic Versions of the EM Algorithm

Abstract

We compare three different stochastic versions of the EM algorithm the SEM algorithm, the SAEM algorithm and the MCEM algorithm. We suggest that the most relevant contribution of the MCEM methodology is what we call the simulated annealing MCEM algorithm, which turns out to be very close to SAEM. We focus particularly on the mixture of distributions problem. In this context, we review the available theoretical results on the convergence of these algorithms and on the behavior of SEM as the sample size tends to infinity. Finally, we illustrated these results with some Monte-Carlo numerical simulation.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1992
Accession Number
ADA246929

Entities

People

  • Gilles Celeux
  • Jean Diebolt
  • Jean-claude Biscarat

Organizations

  • University of Washington

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Computational Science
  • Computations
  • Convergence
  • Data Analysis
  • Data Science
  • Information Science
  • Markov Chains
  • Monte Carlo Method
  • New York
  • Probability
  • Probability Distributions
  • Simulations
  • Statistical Algorithms
  • Statistical Analysis
  • Statistical Inference
  • Statistics

Readers

  • Computational Fluid Dynamics (CFD)
  • Operations Research