Maximum likelihood estimation in discrete mixed hidden Markov models using the SAEM algorithm

摘要: Mixed hidden Markov models have been recently defined in the literature as an extension of for dealing with population studies. The notion mixed is particularly relevant modeling longitudinal data collected during clinical trials, especially when distinct disease stages can be considered. However, parameter estimation such complex, due to their highly nonlinear structure and presence unobserved states. Moreover, existing inference algorithms are extremely time consuming model includes several random effects. New procedures proposed estimating parameters, individual parameters sequences states models. main contribution consists a specific version stochastic approximation EM algorithm coupled Baum-Welch parameters. properties this investigated via Monte-Carlo simulation study, application description daily seizure counts epileptic patients presented.