Fixed-Form Variational Posterior Approximation through Stochastic Linear Regression

2012/06/30 by Tim Salimans, David A. Knowles · 2 citations
Mathematics · Computer Science · #stat.CO #cs.CV #stat.ML #msc:62F15

paper · pdf · doi:10.1214/13-ba858

published as Bayesian Analysis, Volume 8, Number 4 (2013), 837-882

arxiv created 2014/07/28 · arxiv updated 2014/07/29

Abstract

We propose a general algorithm for approximating nonstandard Bayesian posterior distributions. The algorithm minimizes the Kullback-Leibler divergence of an approximating distribution to the intractable posterior distribution. Our method can be used to approximate any posterior distribution, provided that it is given in closed form up to the proportionality constant. The approximation can be any distribution in the exponential family or any mixture of such distributions, which means that it can be made arbitrarily precise. Several examples illustrate the speed and accuracy of our approximation method in practice.

Cited by

• Using Variational Inference to Improve the Efficiency of MCMC Algorithms
• bayesics: Core Statistical Methods via Bayesian Inference in R