Accelerating MCMC with Parallel Predictive Prefetching
![Thumbnail](/bitstream/handle/1/22772115/87215436.pdf.jpg?sequence=4&isAllowed=y)
View/ Open
Published Version
http://auai.org/uai2014/proceedings/uai-2014-proceedings.pdfMetadata
Show full item recordCitation
Angelino, Elaine, Eddie Kohler, Amos Waterland, Margo Seltzer, and Ryan P. Adams. 2014. "Accelerating MCMC with Parallel Predictive Prefetching." In Proceedings of the 30th Conference on Uncertainty in Artificial Intelligence, Quebec City, Canada, July 23-27, 2014: 22-31.Abstract
Parallel predictive prefetching is a new frame- work for accelerating a large class of widely- used Markov chain Monte Carlo (MCMC) algorithms. It speculatively evaluates many potential steps of an MCMC chain in parallel while exploiting fast, iterative approximations to the target density. This can accelerate sampling from target distributions in Bayesian inference problems. Our approach takes advantage of whatever parallel resources are available, but produces results exactly equivalent to standard serial execution. In the initial burn-in phase of chain evaluation, we achieve speedup close to linear in the number of available cores.Terms of Use
This article is made available under the terms and conditions applicable to Open Access Policy Articles, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#OAPCitable link to this page
http://nrs.harvard.edu/urn-3:HUL.InstRepos:22772115
Collections
- FAS Scholarly Articles [18304]
Contact administrator regarding this item (to report mistakes or request changes)