A0616
Title: Bayesian analysis of hidden Markov varying coefficient models with zero-effect regions
Authors: Hefei Liu - Qujing Normal University (China) [presenting]
Abstract: In psychological, social, behavioral, and medical studies, hidden Markov models have been extensively applied to the simultaneous modeling of heterogeneous observation and hidden transition in longitudinal data analysis. However, the existing hidden Markov models are developed in a constant-coefficient framework. A novel hidden Markov varying coefficient model is considered, in which the dynamic relations in the effects of potential covariates on the response variable can be investigated. A Bayesian inference method is developed to estimate the model, especially to detect the zero-effect regions. The empirical performance of the proposed method is evaluated through simulation studies.
