Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 99 - Applied Bayesian Methods in Sciences
Type: Topic Contributed
Date/Time: Monday, August 8, 2022 : 8:30 AM to 10:20 AM
Sponsor: Korean International Statistical Society
Abstract #323586
Title: Bayesian Variable Selection in Generalized Linear Mixed Models Using Gaussian and Diffused-Gamma Prior with Application to Breastfeeding Study
Author(s): Jiyeon Song* and Elizabeth Schifano and Dipak K Dey
Companies: University of Michigan and University of Connecticut and University of Connecticut
Keywords: Bayesian model selection; Mixed effect models; Markov chain Monte Carlo; Variance component

We focus on the problem of identifying significant variables in high-dimensional repeated measures and longitudinal data. In this article, we introduce a Bayesian approach to select associated fixed and random variables in generalized linear regression models. The Gaussian and diffused-gamma prior for fixed effects and other default priors for random effects are proposed, and an iterated conditional modes algorithm and Markov chain Monte Carlo algorithm are developed in this article. To demonstrate the applicability of the proposed method, we provide simulation studies for different types of responses following Gaussian, binomial, and Poisson distributions. The proposed methodology is further applied to a breastfeeding dataset with the analysis of nipple and breast pain severity and exclusive breastfeeding status.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program