Optimization via simulation for maximum likelihood estimation in incomplete data models

Author

Fort, Gersende ; Cappé, Olivier ; Moulines, Eric ; Soulier, Philippe

Author_Institution

ENST, CNRS, Paris, France

fYear

1998

fDate

14-16 Sep 1998

Firstpage

80

Lastpage

83

Abstract

Optimization via simulation is a promising approach for solving maximum likelihood problems in incomplete data models. Among the techniques proposed to date, the Monte-Carlo EM algorithm (MCEM) proposed by Wei and Tanner (1991) has a strong potential but very little is known on its behavior and on strategies for monitoring its convergence. In this contribution, the convergence of MCEM is investigated with a particular emphasis on the stability issue (which is not guaranteed in the original algorithm described by Wei and Tanner). A random truncation strategy, inspired by the Chen´s truncation method for stochastic approximation algorithms, is proposed and analyzed. Finally, the application of our results to blind estimation problems in which the complete data likelihood is from the exponential family is discussed

Keywords

Monte Carlo methods; convergence of numerical methods; digital simulation; maximum likelihood estimation; numerical stability; optimisation; parameter estimation; signal processing; stochastic processes; Chen truncation method; Monte-Carlo EM algorithm; algorithm stability; blind estimation problems; convergence; incomplete data models; maximum likelihood estimation; optimization; random truncation strategy; signal processing; simulation; stochastic approximation algorithms; Convergence; Data models; Iterative algorithms; Markov processes; Maximum likelihood estimation; Random variables; Stability; Stochastic processes;

fLanguage

English

Publisher

ieee

Conference_Titel

Statistical Signal and Array Processing, 1998. Proceedings., Ninth IEEE SP Workshop on

Conference_Location

Portland, OR

Print_ISBN

0-7803-5010-3

Type

conf

DOI

10.1109/SSAP.1998.739339

Filename

739339