Afficher la notice abrégée

dc.contributor.authorTitterington, Mike
dc.contributor.authorRobert, Christian P.
dc.contributor.authorMarin, Jean-Michel
dc.contributor.authorCucala, Lionel
dc.date.accessioned2009-07-07T09:37:48Z
dc.date.available2009-07-07T09:37:48Z
dc.date.issued2009
dc.identifier.urihttps://basepub.dauphine.fr/handle/123456789/867
dc.language.isoenen
dc.subjectBayesian inferenceen
dc.subjectBoltzmann modelen
dc.subjectCompatible conditionalsen
dc.subjectClassificationen
dc.subjectMCMC algorithmsen
dc.subjectPseudo-likelihooden
dc.subjectPath samplingen
dc.subjectPerfect samplingen
dc.subjectNormalising constanten
dc.subject.ddc519en
dc.titleA Bayesian reassessment of nearest-neighbour classificationen
dc.typeArticle accepté pour publication ou publié
dc.contributor.editoruniversityotherUniversity of Glasgow;Royaume-Uni
dc.contributor.editoruniversityotherINRIA – Université Paris Sud - Paris XI;France
dc.contributor.editoruniversityotherINSEE;France
dc.description.abstractenThe k-nearest-neighbour procedure is a well-known deterministic method used in supervised classification. This paper proposes a reassessment of this approach as a statistical technique derived from a proper probabilistic model; in particular, we modify the assessment made in a previous analysis of this method undertaken by Holmes and Adams (2002, 2003), and evaluated by Manocha and Girolami (2007), where the underlying probabilistic model is not completely well-defined. Once a clear probabilistic basis for the k-nearest-neighbour procedure is established, we derive computational tools for conducting Bayesian inference on the parameters of the corresponding model. In particular, we assess the difficulties inherent to pseudo-likelihood and to path sampling approximations of an intractable normalising constant, and propose a perfect sampling strategy to implement a correct MCMC sampler associated with our model. If perfect sampling is not available, we suggest using a Gibbs sampling approximation. Illustrations of the performance of the corresponding Bayesian classifier are provided for several benchmark datasets, demonstrating in particular the limitations of the pseudo-likelihood approximation in this set-up.en
dc.relation.isversionofjnlnameJournal of the American Statistical Association
dc.relation.isversionofjnlvol104
dc.relation.isversionofjnlissue485
dc.relation.isversionofjnldate2009-03
dc.relation.isversionofjnlpages263-273
dc.relation.isversionofdoihttp://dx.doi.org/10.1198/jasa.2009.0125
dc.identifier.urlsitehttp://hal.inria.fr/inria-00143783/en/en
dc.description.sponsorshipprivateouien
dc.relation.isversionofjnlpublisherAmerican Statistical Association
dc.subject.ddclabelProbabilités et mathématiques appliquéesen


Fichiers attachés à cette notice

FichiersTailleFormatConsulter

Il n'y a pas de fichiers associés à cette notice.

Ce document fait partie de la (des) collection(s) suivante(s)

Afficher la notice abrégée