Show simple item record

dc.contributor.authorBurt, Daviden
dc.contributor.authorRasmussen, Carlen
dc.contributor.authorVan Der Wilk, Men
dc.date.accessioned2020-08-05T23:30:26Z
dc.date.available2020-08-05T23:30:26Z
dc.date.issued2020-07-01en
dc.identifier.issn1532-4435
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/308821
dc.description.abstractGaussian processes are distributions over functions that are versatile and mathematically convenient priors in Bayesian modelling. However, their use is often impeded for data with large numbers of observations, N, due to the cubic (in N) cost of matrix operations used in exact inference. Many solutions have been proposed that rely on M << N inducing variables to form an approximation at a cost of O(NM^2). While the computational cost appears linear in N, the true complexity depends on how M must scale with N to ensure a certain quality of the approximation. In this work, we investigate upper and lower bounds on how M needs to grow with N to ensure high quality approximations. We show that we can make the KL-divergence between the approximate model and the exact posterior arbitrarily small for a Gaussian-noise regression model with M<<N. Specifically, for the popular squared exponential kernel and D-dimensional Gaussian distributed covariates, M=O((log N)^D) suffice and a method with an overall computational cost of O(N(log N)^{2D}(\log\log N)^2) can be used to perform inference.
dc.publisherMicrotome Publishing
dc.rightsAttribution 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.titleConvergence of sparse variational inference in gaussian processes regressionen
dc.typeArticle
prism.publicationDate2020en
prism.publicationNameJournal of Machine Learning Researchen
prism.volume21en
dc.identifier.doi10.17863/CAM.55909
dcterms.dateAccepted2020-03-26en
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://creativecommons.org/licenses/by/4.0/en
rioxxterms.licenseref.startdate2020-07-01en
dc.contributor.orcidRasmussen, Carl [0000-0001-8899-7850]
dc.identifier.eissn1533-7928
rioxxterms.typeJournal Article/Reviewen


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution 4.0 International
Except where otherwise noted, this item's licence is described as Attribution 4.0 International