Show simple item record

dc.contributor.authorFrellsen, Jesen
dc.contributor.authorWinther, Oleen
dc.contributor.authorGhahramani, Zoubinen
dc.contributor.authorFerkinghoff-Borg, Jesperen
dc.date.accessioned2016-03-15T14:48:21Z
dc.date.available2016-03-15T14:48:21Z
dc.date.issued2016en
dc.identifier.citationFrellsen et al. Proceedings of the Nineteenth International Conference on Artificial Intelligence and Statistics (2016)en
dc.identifier.issn1938-7288
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/254509
dc.description.abstractBayesian generalised ensemble (BayesGE) is a new method that addresses two major drawbacks of standard Markov chain Monte Carlo algorithms for inference in high-dimensional probability models: inapplicability to estimate the partition function, and poor mixing properties. BayesGE uses a Bayesian approach to iteratively update the belief about the density of states (distribution of the log likelihood under the prior) for the model, with the dual purpose of enhancing the sampling efficiency and make the estimation of the partition function tractable. We benchmark BayesGE on Ising and Potts systems and show that it compares favourably to existing state-of-the-art methods.
dc.description.sponsorshipJF acknowledge funding from the Danish Council for Independent Research | Natural Sciences. ZG acknowledge funding from EPSRC EP/I036575/1 and Google.
dc.languageEnglishen
dc.language.isoenen
dc.publisherMicrotome Publishing
dc.titleBayesian generalised ensemble Markov chain Monte Carloen
dc.typeArticle
dc.description.versionThis is the author accepted manuscript. It is currently under an indefinite embargo pending publication by Microtome Publishing.en
prism.publicationDate2016en
prism.publicationNameProceedings of the Nineteenth International Conference on Artificial Intelligence and Statisticsen
dc.rioxxterms.funderEPSRC
dc.rioxxterms.projectidEP/I036575/1
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserveden
rioxxterms.licenseref.startdate2016en
dc.contributor.orcidGhahramani, Zoubin [0000-0002-7464-6475]
rioxxterms.typeJournal Article/Reviewen


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record