Constraints in Particle Swarm Optimization of Hidden Markov Models
This paper presents new application of Particle Swarm Optimization (PSO) algorithm for training Hidden Markov Models (HMMs). The problem of finding an optimal set of model parameters is numerical optimization problem constrained by stochastic character of HMM parameters. Constraint handling is carried out using three different ways and the results are compared to Baum-Welch algorithm (BW), commonly used for HMM training. The global searching PSO method is much less sensitive to local extremes and finds better solutions than the local BW algorithm, which often converges to local optima. The advantage of PSO approach was markedly evident, when longer training sequence was used.
KeywordsParticle Swarm Optimization Hide Markov Model Particle Swarm Optimization Algorithm Observation Sequence Constraint Handling
Unable to display preview. Download preview PDF.
- Kwong, S., Chau, C.: Analysis of paralell genetic algortihms on hmm based speech recognition system. In: ICASSP, pp. 1229–1233 (1997)Google Scholar
- Kennedy, J., Eberhart, R.: Particle swarm optimization. In: Proceedings International Conference on Neural Networks IEEE, vol. 4, pp. 1942–1948 (1995)Google Scholar
- Hu, X., Eberhart, R.C.: Solving constrained nonlinear optimization problems with particle swarm optimization. In: Proceedings of the Sixth World Multiconference on Systems, Cybernetics and Informatics (2002)Google Scholar
- Paquet, U., p Engelbrecht, A.: A New Particle Swarm Optimiser for Linearly Constrained Optimization. In: Proceedings of the Congress on Evolutionary Computation 2003 (CEC2003), Piscataway, new Jersey, Canberra, Ausralia, vol. 1, pp. 227–233. IEEE Service Center, Los Alamitos (2003)Google Scholar