Show simple item record

dc.contributor.authorFinke, A
dc.contributor.authorSingh, SS
dc.date.accessioned2017-09-13T13:21:33Z
dc.date.available2017-09-13T13:21:33Z
dc.date.issued2017-08-09
dc.identifier.issn1053-587X
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/267197
dc.description.abstractWe present approximate algorithms for performing smoothing in a class of high-dimensional state-space models via sequential Monte Carlo methods ("particle filters"). In high dimensions, a prohibitively large number of Monte Carlo samples ("particles"), growing exponentially in the dimension of the state space, is usually required to obtain a useful smoother. Employing blocking approximations, we exploit the spatial ergodicity properties of the model to circumvent this curse of dimensionality. We thus obtain approximate smoothers that can be computed recursively in time and parallel in space. First, we show that the bias of our blocked smoother is bounded uniformly in the time horizon and in the model dimension. We then approximate the blocked smoother with particles and derive the asymptotic variance of idealised versions of our blocked particle smoother to show that variance is no longer adversely effected by the dimension of the model. Finally, we employ our method to successfully perform maximum-likelihood estimation via stochastic gradient-ascent and stochastic expectation-maximisation algorithms in a 100-dimensional state-space model.
dc.languageEng
dc.language.isoen
dc.publisherIEEE
dc.subjecthigh dimensions
dc.subjectsmoothing
dc.subjectparticle filter
dc.subjectsequential Monte Carlo
dc.subjectstate-space model
dc.titleApproximate Smoothing and Parameter Estimation in High-Dimensional State-Space Models
dc.typeArticle
prism.issueIdentifier99
prism.publicationDate2017
prism.publicationNameIEEE Transactions on Signal Processing
prism.volumePP
dc.identifier.doi10.17863/CAM.11966
dcterms.dateAccepted2017-07-10
rioxxterms.versionofrecord10.1109/TSP.2017.2733504
rioxxterms.versionAM
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.licenseref.startdate2017-08-09
dc.identifier.eissn1941-0476
dc.publisher.urlhttps://ieeexplore.ieee.org/document/8003455
rioxxterms.typeJournal Article/Review
pubs.funder-project-idEngineering and Physical Sciences Research Council (EP/K020153/1)
cam.issuedOnline2017-08-09
dc.identifier.urlhttps://ieeexplore.ieee.org/document/8003455


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record