Asynchronous Gibbs Sampling

Alexander Terenin, Daniel Simpson, David Draper

Research output: Chapter in Book/Report/Conference proceedingConference PaperOtherpeer-review


Gibbs sampling is a Markov Chain Monte Carlo (MCMC) method often used in Bayesian learning. MCMC methods can be difficult to deploy on parallel and distributed systems due to their inherently sequential nature. We study asynchronous Gibbs sampling, which achieves parallelism by simply ignoring sequential requirements. This method has been shown to produce good empirical results for some hierarchical models, and is popular in the topic modeling community, but was also shown to diverge for other targets. We introduce a theoretical framework for analyzing asynchronous Gibbs sampling and other extensions of MCMC that do not possess the Markov property. We prove that asynchronous Gibbs can be modified so that it converges under appropriate regularity conditions - we call this the exact asynchronous Gibbs algorithm. We study asynchronous Gibbs on a set of examples by comparing the exact and approximate algorithms, including two where it works well, and one where it fails dramatically. We conclude with a set of heuristics to describe settings where the algorithm can be effectively used.
Original languageEnglish
Title of host publicationProceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics
EditorsNeil Lawrence, Mark Reid
Place of PublicationUSA
PublisherSociety for Artificial Intelligence and Statistics
Number of pages11
Publication statusPublished - 2020
Externally publishedYes
EventInternational Conference on Artificial Intelligence and Statistics 2020 - Virtual, Italy
Duration: 3 Jun 20205 Jun 2020
Conference number: 23rd (Website) (Proceedings)

Publication series

NameProceedings of Machine Learning Research
ISSN (Electronic)2640-3498


ConferenceInternational Conference on Artificial Intelligence and Statistics 2020
Abbreviated titleAISTATS 2020
Internet address

Cite this