Repository logo
 

Revisiting Context-Tree Weighting for Bayesian Inference

Simple item page
cam.depositDate2021-12-22
cam.orpheus.successTue Feb 01 19:02:34 GMT 2022 - Embargo updated
dc.contributor.authorPapageorgiou, I
dc.contributor.authorKontoyiannis, I
dc.contributor.authorMertzanis, L
dc.contributor.authorPanotopoulou, A
dc.contributor.authorSkoularidou, M
dc.contributor.orcidKontoyiannis, Ioannis [0000-0001-7242-6375]
dc.date.accessioned2022-01-25T00:30:12Z
dc.date.available2022-01-25T00:30:12Z
dc.date.issued2021
dc.date.updated2021-12-22T14:01:01Z
dc.description.abstractWe revisit the statistical foundation of the celebrated context tree weighting (CTW) algorithm, and we develop a Bayesian modelling framework for the class of higher-order, variable-memory Markov chains, along with an associated collection of methodological tools for exact inference for discrete time series. In addition to deterministic algorithms that learn the a posteriori most likely models and compute their posterior probabilities, we introduce a family of variable-dimension Markov chain Monte Carlo samplers, facilitating further exploration of the posterior. The performance of the proposed methods in model selection, Markov order estimation and prediction is illustrated through simulation experiments and real-world applications.
dc.identifier.doi10.17863/CAM.80321
dc.identifier.isbn9781538682098
dc.identifier.issn2157-8095
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/332891
dc.language.isoeng
dc.publisherIEEE
dc.publisher.departmentDepartment of Pure Mathematics And Mathematical Statistics
dc.publisher.urlhttp://dx.doi.org/10.1109/isit45174.2021.9518189
dc.rightsAll Rights Reserved
dc.rights.urihttp://www.rioxx.net/licenses/all-rights-reserved
dc.subject49 Mathematical Sciences
dc.subject46 Information and Computing Sciences
dc.subject4905 Statistics
dc.subject4611 Machine Learning
dc.subject2.5 Research design and methodologies (aetiology)
dc.subject2 Aetiology
dc.titleRevisiting Context-Tree Weighting for Bayesian Inference
dc.typeConference Object
prism.endingPage2911
prism.publicationDate2021
prism.publicationNameIEEE International Symposium on Information Theory - Proceedings
prism.startingPage2906
prism.volume2021-July
pubs.conference-finish-date2021-07-20
pubs.conference-name2021 IEEE International Symposium on Information Theory (ISIT)
pubs.conference-start-date2021-07-12
pubs.licence-display-nameApollo Repository Deposit Licence Agreement
pubs.licence-identifierapollo-deposit-licence-2-1
rioxxterms.versionAM
rioxxterms.versionofrecord10.1109/ISIT45174.2021.9518189

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
isit_msp.pdf
Size:
431.1 KB
Format:
Adobe Portable Document Format
Description:
Accepted version
Licence
http://www.rioxx.net/licenses/all-rights-reserved
Download

Collections

Cambridge University Research Outputs