Particle methods for maximum likelihood estimation in latent variable models
- 297 Downloads
Standard methods for maximum likelihood parameter estimation in latent variable models rely on the Expectation-Maximization algorithm and its Monte Carlo variants. Our approach is different and motivated by similar considerations to simulated annealing; that is we build a sequence of artificial distributions whose support concentrates itself on the set of maximum likelihood estimates. We sample from these distributions using a sequential Monte Carlo approach. We demonstrate state-of-the-art performance for several applications of the proposed approach.
KeywordsLatent variable models Markov chain Monte Carlo Maximum likelihood Sequential Monte Carlo Simulated annealing
Unable to display preview. Download preview PDF.
- Johansen, A.M.: Some non-standard sequential Monte Carlo methods with applications, Ph.D. thesis. University of Cambridge, Department of Engineering (2006) Google Scholar