Mean Field Variational Bayes for Finite Mixture of Random Coefficients Models

For over three decades, random coefficient models, especially for panel data, have been prevalent. Categorical random coefficient models relate to finite mixtures of normal regressions. The talk presents Markov Chain Monte Carlo (MCMC) and mean field variational Bayes (MFVB) methods for finite mixtures of random coefficients models.