EM algorithm and varible Neighborhood Search for fitting Finite Mixture Model parameters

Author

Bessadok, A. ; Hansen, Paul ; Rebai, Abdelwaheb

Author_Institution

Inst. Super. de Gestion, Univ. de GABES, Tunisia

fYear

2009

fDate

12-14 Oct. 2009

Firstpage

725

Lastpage

733

Abstract

Finding maximum likelihood parameter values for Finite Mixture Model (FMM) is often done with the Expectation Maximization (EM) algorithm. However the choice of initial values can severely affect the time to attain convergence of the algorithm and its efficiency in finding global maxima. We alleviate this defect by embedding the EM algorithm within the variable Neighborhood Search (VNS) methaheurestic framework. Computational experiment in several problems in literature as well as some larger ones are reported.

Keywords

Gaussian distribution; expectation-maximisation algorithm; optimisation; expectation maximization algorithm; finite mixture model parameters; global maxima; maximum likelihood parameter; variable neighborhood search; Biological system modeling; Computer science; Convergence; Electronic mail; Equations; Information technology; Iterative algorithms; Iterative methods; Maximum likelihood estimation; Parameter estimation; Expectation Maximization algorithm; Finite Gaussian Mixture Model and Global Optimization‥; Maximum Likelihood Estimation; Metaheuristic; Variable Neighborhood Search;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Information Technology, 2009. IMCSIT '09. International Multiconference on

Conference_Location

Mragowo

Print_ISBN

978-1-4244-5314-6

Type

conf

DOI

10.1109/IMCSIT.2009.5352758

Filename

5352758