Integrated non-factorized variational inference

Shaobo Han, Xuejun Liao, Lawrence Carin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

We present a non-factorized variational method for full posterior inference in Bayesian hierarchical models, with the goal of capturing the posterior variable dependencies via efficient and possibly parallel computation. Our approach unifies the integrated nested Laplace approximation (INLA) under the variational framework. The proposed method is applicable in more challenging scenarios than typically assumed by INLA, such as Bayesian Lasso, which is characterized by the non-differentiability of the ℓ1 norm arising from independent Laplace priors. We derive an upper bound for the Kullback-Leibler divergence, which yields a fast closed-form solution via decoupled optimization. Our method is a reliable analytic alternative to Markov chain Monte Carlo (MCMC), and it results in a tighter evidence lower bound than that of mean-field variational Bayes (VB) method.
Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
StatePublished - Jan 1 2013
Externally publishedYes

Fingerprint Dive into the research topics of 'Integrated non-factorized variational inference'. Together they form a unique fingerprint.

Cite this