Please use this identifier to cite or link to this item: https://doi.org/10.1214/13-STS418
DC FieldValue
dc.titleVariational inference for generalized linear mixed models using partially noncentered parametrizations
dc.contributor.authorTan, L.S.L.
dc.contributor.authorNott, D.J.
dc.date.accessioned2014-10-28T05:16:30Z
dc.date.available2014-10-28T05:16:30Z
dc.date.issued2013-05
dc.identifier.citationTan, L.S.L., Nott, D.J. (2013-05). Variational inference for generalized linear mixed models using partially noncentered parametrizations. Statistical Science 28 (2) : 168-188. ScholarBank@NUS Repository. https://doi.org/10.1214/13-STS418
dc.identifier.issn08834237
dc.identifier.urihttp://scholarbank.nus.edu.sg/handle/10635/105461
dc.description.abstractThe effects of different parametrizations on the convergence of Bayesian computational algorithms for hierarchical models are well explored. Techniques such as centering, noncentering and partial noncentering can be used to accelerate convergence in MCMC and EM algorithms but are still not well studied for variational Bayes (VB) methods. As a fast deterministic approach to posterior approximation, VB is attracting increasing interest due to its suitability for large high-dimensional data. Use of different parametrizations for VB has not only computational but also statistical implications, as different parametrizations are associated with different factorized posterior approximations. We examine the use of partially noncentered parametrizations in VB for generalized linear mixed models (GLMMs). Our paper makes four contributions. First, we show how to implement an algorithm called nonconjugate variational message passing for GLMMs. Second, we show that the partially noncentered parametrization can adapt to the quantity of information in the data and determine a parametrization close to optimal. Third, we show that partial noncentering can accelerate convergence and produce more accurate posterior approximations than centering or noncentering. Finally, we demonstrate how the variational lower bound, produced as part of the computation, can be useful for model selection. © Institute of Mathematical Statistics, 2013.
dc.description.urihttp://libproxy1.nus.edu.sg/login?url=http://dx.doi.org/10.1214/13-STS418
dc.sourceScopus
dc.subjectHierarchical centering
dc.subjectLongitudinal data analysis.
dc.subjectNonconjugate models
dc.subjectVariational bayes
dc.subjectVariational message passing
dc.typeArticle
dc.contributor.departmentSTATISTICS & APPLIED PROBABILITY
dc.description.doi10.1214/13-STS418
dc.description.sourcetitleStatistical Science
dc.description.volume28
dc.description.issue2
dc.description.page168-188
dc.identifier.isiut000319892300002
Appears in Collections:Staff Publications

Show simple item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

26
checked on Nov 25, 2022

WEB OF SCIENCETM
Citations

24
checked on Nov 25, 2022

Page view(s)

139
checked on Nov 24, 2022

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.