Skip to main content

Bayesian Variable Selection for Latent Class Models

Date:
-
Location:
University of Kentucky, Whitehall Classroom Building room 102
Speaker(s) / Presenter(s):
Joyee Ghosh, University of North Carolina

 

In this talk, I will describe a latent class model with class probabilities that depend on subjectspecific covariates. Our methods are motivated primarily by an application in reproductive epidemiology, where it is of interest to identify important predictors of latent pregnancy weight gain classes. We consider methodology that allows estimation of latent classes while allowing for variable selection uncertainty. We propose a Bayesian variable selection approach and implement a stochastic search Gibbs sampler for posterior computation to obtain model averaged estimates of quantities of interest such as marginal inclusion probabilities of predictors. We evaluate the performance of our methodology through simulation studies and apply it to data from the motivating application. 

Event Series: