Show simple item record

dc.contributor.authorWang, Y
dc.contributor.authorWipf, D
dc.contributor.authorLing, Q
dc.contributor.authorChen, W
dc.contributor.authorWassell, I
dc.date.accessioned2015-10-02T14:12:13Z
dc.date.available2015-10-02T14:12:13Z
dc.date.issued2015
dc.identifier.citationInternational Conference on Machine Learning 2015, 1209-1217.
dc.identifier.isbn9781510810587
dc.identifier.issn2640-3498
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/251288
dc.description.abstractSubspace segmentation is the process of clustering a set of data points that are assumed to lie on the union of multiple linear or affine subspaces, and is increasingly being recognized as a fundamental tool for data analysis in high dimensional settings. Arguably one of the most successful approaches is based on the observation that the sparsest representation of a given point with respect to a dictionary formed by the others involves nonzero coefficients associated with points originating in the same subspace. Such sparse representations are computed independently for each data point via ℓ1-norm minimization and then combined into an affinity matrix for use by a final spectral clustering step. The downside of this procedure is two-fold. First, unlike canonical compressive sensing scenarios with ideally-randomized dictionaries, the data-dependent dictionaries here are unavoidably highly structured, disrupting many of the favorable properties of the ℓ1 norm. Secondly, by treating each data point independently, we ignore useful relationships between points that can be leveraged for jointly computing such sparse representations. Consequently, we motivate a multi-task learning-based framework for learning coupled sparse representations leading to a segmentation pipeline that is both robust against correlation structure and tailored to generate an optimal affinity matrix. Theoretical analysis and empirical tests are provided to support these claims.
dc.description.sponsorshipY. Wang is sponsored by the University of Cambridge Overseas Trust. Y. Wang and Q. Ling are partially supported by sponsorship from Microsoft Research Asia. Q. Ling is also supported in part by NSFC grant 61004137. W. Chen is supported by EPSRC Research Grant EP/K033700/1 and the Natural Science Foundation of China 61401018.
dc.languageEnglish
dc.language.isoen
dc.publisherJMLR.org
dc.rightsAttribution-NonCommercial 2.0 UK: England & Wales
dc.rights.urihttp://creativecommons.org/licenses/by-nc/2.0/uk/
dc.titleMulti-task learing for subspace segmentation
dc.typeConference Object
dc.description.versionThis is the final version of the article. It first appeared from JMLR via http://jmlr.org/proceedings/papers/v37/wangc15.html
prism.endingPage1217
prism.publicationDate2015
prism.publicationName32nd International Conference on Machine Learning, ICML 2015
prism.startingPage1209
prism.volume32
dc.rioxxterms.funderEPSRC
dc.rioxxterms.projectidEP/K033700/1
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.licenseref.startdate2015
dc.contributor.orcidWassell, Ian [0000-0001-7927-5565]
dc.publisher.urlhttp://proceedings.mlr.press/v37/
rioxxterms.typeConference Paper/Proceeding/Abstract
pubs.conference-nameInternational Conference on Machine Learning 2015 (ICML 2015)


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial 2.0 UK: England & Wales
Except where otherwise noted, this item's licence is described as Attribution-NonCommercial 2.0 UK: England & Wales