Variational Bayes for Elaborate Distributions

Authors

M. Wand, University of WollongongFollow
J. T. Ormerod, University of Sydney
S. A. Padoan, Swiss Federal Institute of Technology, Lausanne
R. Fruhwirth, Austrian Academy of Sciences, Vienna

Publication Date

2010

Recommended Citation

Wand, M.; Ormerod, J. T.; Padoan, S. A.; and Fruhwirth, R., Variational Bayes for Elaborate Distributions, Centre for Statistical and Survey Methodology, University of Wollongong, Working Paper 05-10, 2010, 35p.
https://ro.uow.edu.au/cssmwp/56

Abstract

We develop strategies for variational Bayes approximate inference for models containing elaborate distributions. Such models suffer from the difficulty that the parameter updates do not admit closed form solutions. We circumvent this problem through a combination of (a) specially tailored auxiliary variables, (b) univariate quadrature schemes and (c) finite mixture approximations of troublesome density functions. An accuracy assessment is conducted and the new methodology is illustrated in an application.