Title: Convergence of position-dependent MALA with application to conditional simulation in GLMMs
Authors: Vivekananda Roy - Iowa State University (United States) [presenting]
Abstract: After discussing different variants of the Metropolis adjusted Langevin algorithms (MALA), we describe some convergence results for these Markov chains. The likelihood function in generalized linear mixed models (GLMMs) is available only as a high dimensional integral, and thus the resulting posterior densities in GLMMs are intractable. We study and compare the performance of variants of MALA in the context of conditional simulation from the two most popular GLMMs, namely the binomial GLMM with logit link and the Poisson GLMM with log link.