File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: CAST: A Correlation-based Adaptive Spectral Clustering Algorithm on Multi-scale Data

TitleCAST: A Correlation-based Adaptive Spectral Clustering Algorithm on Multi-scale Data
Authors
Keywordsmulti-scale data
robustness
spectral clustering
Issue Date2020
PublisherAssociation for Computing Machinery.
Citation
Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual Conference, San Diego, CA, USA, 22-27 August 2020, p. 439-449 How to Cite?
AbstractWe study the problem of applying spectral clustering to cluster multi-scale data, which is data whose clusters are of various sizes and densities. Traditional spectral clustering techniques discover clusters by processing a similarity matrix that reflects the proximity of objects. For multi-scale data, distance-based similarity is not effective because objects of a sparse cluster could be far apart while those of a dense cluster have to be sufficiently close. Following [16], we solve the problem of spectral clustering on multi-scale data by integrating the concept of objects' 'reachability similarity' with a given distance-based similarity to derive an objects' coefficient matrix. We propose the algorithm CAST that applies trace Lasso to regularize the coefficient matrix. We prove that the resulting coefficient matrix has the 'grouping effect' and that it exhibits 'sparsity'. We show that these two characteristics imply very effective spectral clustering. We evaluate CAST and 10 other clustering methods on a wide range of datasets w.r.t. various measures. Experimental results show that CAST provides excellent performance and is highly robust across test cases of multi-scale data.
DescriptionSession: Research Track Papers
Persistent Identifierhttp://hdl.handle.net/10722/289848
ISBN

 

DC FieldValueLanguage
dc.contributor.authorLi, X-
dc.contributor.authorKao, CM-
dc.contributor.authorShan, C-
dc.contributor.authorYin, D-
dc.contributor.authorEster, M-
dc.date.accessioned2020-10-22T08:18:20Z-
dc.date.available2020-10-22T08:18:20Z-
dc.date.issued2020-
dc.identifier.citationProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual Conference, San Diego, CA, USA, 22-27 August 2020, p. 439-449-
dc.identifier.isbn9781450379984-
dc.identifier.urihttp://hdl.handle.net/10722/289848-
dc.descriptionSession: Research Track Papers-
dc.description.abstractWe study the problem of applying spectral clustering to cluster multi-scale data, which is data whose clusters are of various sizes and densities. Traditional spectral clustering techniques discover clusters by processing a similarity matrix that reflects the proximity of objects. For multi-scale data, distance-based similarity is not effective because objects of a sparse cluster could be far apart while those of a dense cluster have to be sufficiently close. Following [16], we solve the problem of spectral clustering on multi-scale data by integrating the concept of objects' 'reachability similarity' with a given distance-based similarity to derive an objects' coefficient matrix. We propose the algorithm CAST that applies trace Lasso to regularize the coefficient matrix. We prove that the resulting coefficient matrix has the 'grouping effect' and that it exhibits 'sparsity'. We show that these two characteristics imply very effective spectral clustering. We evaluate CAST and 10 other clustering methods on a wide range of datasets w.r.t. various measures. Experimental results show that CAST provides excellent performance and is highly robust across test cases of multi-scale data.-
dc.languageeng-
dc.publisherAssociation for Computing Machinery.-
dc.relation.ispartofProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining-
dc.rightsProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. Copyright © Association for Computing Machinery.-
dc.subjectmulti-scale data-
dc.subjectrobustness-
dc.subjectspectral clustering-
dc.titleCAST: A Correlation-based Adaptive Spectral Clustering Algorithm on Multi-scale Data-
dc.typeConference_Paper-
dc.identifier.emailLi, X: xli2@hku.hk-
dc.identifier.emailKao, CM: kao@cs.hku.hk-
dc.identifier.authorityKao, CM=rp00123-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1145/3394486.3403086-
dc.identifier.scopuseid_2-s2.0-85090426592-
dc.identifier.hkuros316381-
dc.identifier.spage439-
dc.identifier.epage449-
dc.publisher.placeNew York, NY-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats