Title: A low complexity simulated annealing approach for training hidden Markov models

Authors: Tarik Al-Ani, Yskandar Hamam

Addresses: Laboratoire d'Ingenierie des Systemes de Versailles (LISV), Universite de Versailles Saint Quentin, 10-12 Avenue de l. Europe, Velizy 78140, France; Departement Informatique, ESIEE-PARIS, Cite Descartes BP 99 93162, Noisy-Le-Grand, France. ' F'SATIE, Tshwane University of Technology, Private Bag X680, Pretoria 0001, South Africa; Departement Informatique, ESIEE-PARIS, Cite Descartes BP 99 93162, Noisy-Le-Grand, France; Laboratoire d'Ingenierie des Systemes de Versailles (LISV), Universite de Versailles Saint Quentin, 10-12 Avenue de l.Europe, Velizy 78140, France

Abstract: An algorithm for the training of Hidden Markov Models (HMMs) by simulated annealing is presented. This algorithm is based on a finite coding of the solution space based on the optimal trajectory of the state. It is applied to both discrete and continuous Gaussian observations. The algorithm needs no specific initialisation of the initial HMM by the user, the cooling schedule being general and applicable to any specific model. The parameters of the algorithm (initial and final temperatures) are derived automatically from theoretical considerations. The objective function evaluations of the algorithm are made independent of the problem size in order to minimise the computation time. A comparative study between the conventional Baum–Welch algorithm, Viterbi based algorithm and our simulated annealing algorithm shows that our algorithm gives better results and overcome the problem of HMM initialisation needed by the others.

Keywords: hidden Markov models; HMM; operational research; simulated annealing; finite coding; solution space; optimal trajectory.

DOI: 10.1504/IJOR.2010.034070

International Journal of Operational Research, 2010 Vol.8 No.4, pp.483 - 510

Published online: 07 Jul 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article