Using Reversible Jump MCMC To Solve Latent Class Analysis

Latent class model is a type of latent variable model,which is widely used inpsychology and social science.It is called a latent class model because the latentvariable is discrete.Traditionally we estimate the number of components and themixture component parameters separately,but in this paper we employ reversiblejump Markov Chain Monte Carlo(rjMCMC)method to estimate the number ofcomponents and component parameters jointly on the basis of Bayesian theory,itis better than traditional method to solve latent class analysis.