liu.seSök publikationer i DiVA
Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Speeding up MCMC by Delayed Acceptance and Data Subsampling
Linköpings universitet, Institutionen för datavetenskap, Statistik. Linköpings universitet, Tekniska fakulteten. Research Division, Sveriges Riksbank, Stockholm, Sweden.
Discipline of Business Analytics, University of Sydney, Camperdown NSW, Australia.
Linköpings universitet, Institutionen för datavetenskap, Statistik och maskininlärning. Linköpings universitet, Tekniska fakulteten. (Division of Statistics and Machine Learning (STIMA))ORCID-id: 0000-0003-2786-2519
Australian School of Business, University of New South Wales, Sydney NSW, Australia.
2018 (Engelska)Ingår i: Journal of Computational And Graphical Statistics, ISSN 1061-8600, E-ISSN 1537-2715, Vol. 27, nr 1, s. 12-22Artikel i tidskrift (Refereegranskat) Published
Abstract [en]

The complexity of the Metropolis–Hastings (MH) algorithm arises from the requirement of a likelihood evaluation for the full dataset in each iteration. One solution has been proposed to speed up the algorithm by a delayed acceptance approach where the acceptance decision proceeds in two stages. In the first stage, an estimate of the likelihood based on a random subsample determines if it is likely that the draw will be accepted and, if so, the second stage uses the full data likelihood to decide upon final acceptance. Evaluating the full data likelihood is thus avoided for draws that are unlikely to be accepted. We propose a more precise likelihood estimator that incorporates auxiliary information about the full data likelihood while only operating on a sparse set of the data. We prove that the resulting delayed acceptance MH is more efficient. The caveat of this approach is that the full dataset needs to be evaluated in the second stage. We therefore propose to substitute this evaluation by an estimate and construct a state-dependent approximation thereof to use in the first stage. This results in an algorithm that (i) can use a smaller subsample m by leveraging on recent advances in Pseudo-Marginal MH (PMMH) and (ii) is provably within O(m^-2) of the true posterior.

Ort, förlag, år, upplaga, sidor
Taylor & Francis Group, 2018. Vol. 27, nr 1, s. 12-22
Nyckelord [en]
Bayesian inference, Delayed acceptance MCMC, Large data, Markov chain Monte Carlo, Survey sampling
Nationell ämneskategori
Sannolikhetsteori och statistik
Identifikatorer
URN: urn:nbn:se:liu:diva-140873DOI: 10.1080/10618600.2017.1307117ISI: 000430484300002Scopus ID: 2-s2.0-85026419157OAI: oai:DiVA.org:liu-140873DiVA, id: diva2:1141080
Forskningsfinansiär
Stiftelsen för strategisk forskning (SSF), RIT 15-0097
Anmärkning

Funding agencies: VINNOVA grant [2010-02635]; Business School Pilot Research grant; Swedish Foundation for Strategic Research [RIT 15-0097]; Australian Research Council Centre of Excellence grant [CE140100049]

Tillgänglig från: 2017-09-13 Skapad: 2017-09-13 Senast uppdaterad: 2019-12-30Bibliografiskt granskad

Open Access i DiVA

fulltext(616 kB)560 nedladdningar
Filinformation
Filnamn FULLTEXT01.pdfFilstorlek 616 kBChecksumma SHA-512
705ea423c86904b8f6c2dff15e5157b409ff93a1dcbcabfcf8ca52fdff63c84fbce2f74e91a32bddd89ea4a1bca7e2d6d83aeb0d67de67d48e4f51410cf532e9
Typ fulltextMimetyp application/pdf

Övriga länkar

Förlagets fulltextScopus

Person

Villani, Mattias

Sök vidare i DiVA

Av författaren/redaktören
Quiroz, MatiasVillani, Mattias
Av organisationen
StatistikTekniska fakultetenStatistik och maskininlärning
I samma tidskrift
Journal of Computational And Graphical Statistics
Sannolikhetsteori och statistik

Sök vidare utanför DiVA

GoogleGoogle Scholar
Totalt: 565 nedladdningar
Antalet nedladdningar är summan av nedladdningar för alla fulltexter. Det kan inkludera t.ex tidigare versioner som nu inte längre är tillgängliga.

doi
urn-nbn

Altmetricpoäng

doi
urn-nbn
Totalt: 206 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf