Evidential clustering of large dissimilarity data

© 2016 Elsevier B.V. In evidential clustering, the membership of objects to clusters is considered to be uncertain and is represented by Dempster-Shafer mass functions, forming a credal partition. The EVCLUS algorithm constructs a credal partition in such a way that larger dissimilarities between o...

Full description

Saved in:
Bibliographic Details
Main Authors: Denœux T., Sriboonchitta S., Kanjanatarakul O.
Format: Journal
Published: 2017
Online Access:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84973519329&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/41653
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
id th-cmuir.6653943832-41653
record_format dspace
spelling th-cmuir.6653943832-416532017-09-28T04:22:37Z Evidential clustering of large dissimilarity data Denœux T. Sriboonchitta S. Kanjanatarakul O. © 2016 Elsevier B.V. In evidential clustering, the membership of objects to clusters is considered to be uncertain and is represented by Dempster-Shafer mass functions, forming a credal partition. The EVCLUS algorithm constructs a credal partition in such a way that larger dissimilarities between objects correspond to higher degrees of conflict between the associated mass functions. In this paper, we present several improvements to EVCLUS, making it applicable to very large dissimilarity data. First, the gradient-based optimization procedure in the original EVCLUS algorithm is replaced by a much faster iterative row-wise quadratic programming method. Secondly, we show that EVCLUS can be provided with only a random sample of the dissimilarities, reducing the time and space complexity from quadratic to roughly linear. Finally, we introduce a two-step approach to construct credal partitions assigning masses to selected pairs of clusters, making the algorithm outputs more informative than those of the original EVCLUS, while remaining manageable for large numbers of clusters. 2017-09-28T04:22:37Z 2017-09-28T04:22:37Z 2016-08-15 Journal 09507051 2-s2.0-84973519329 10.1016/j.knosys.2016.05.043 https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84973519329&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/41653
institution Chiang Mai University
building Chiang Mai University Library
country Thailand
collection CMU Intellectual Repository
description © 2016 Elsevier B.V. In evidential clustering, the membership of objects to clusters is considered to be uncertain and is represented by Dempster-Shafer mass functions, forming a credal partition. The EVCLUS algorithm constructs a credal partition in such a way that larger dissimilarities between objects correspond to higher degrees of conflict between the associated mass functions. In this paper, we present several improvements to EVCLUS, making it applicable to very large dissimilarity data. First, the gradient-based optimization procedure in the original EVCLUS algorithm is replaced by a much faster iterative row-wise quadratic programming method. Secondly, we show that EVCLUS can be provided with only a random sample of the dissimilarities, reducing the time and space complexity from quadratic to roughly linear. Finally, we introduce a two-step approach to construct credal partitions assigning masses to selected pairs of clusters, making the algorithm outputs more informative than those of the original EVCLUS, while remaining manageable for large numbers of clusters.
format Journal
author Denœux T.
Sriboonchitta S.
Kanjanatarakul O.
spellingShingle Denœux T.
Sriboonchitta S.
Kanjanatarakul O.
Evidential clustering of large dissimilarity data
author_facet Denœux T.
Sriboonchitta S.
Kanjanatarakul O.
author_sort Denœux T.
title Evidential clustering of large dissimilarity data
title_short Evidential clustering of large dissimilarity data
title_full Evidential clustering of large dissimilarity data
title_fullStr Evidential clustering of large dissimilarity data
title_full_unstemmed Evidential clustering of large dissimilarity data
title_sort evidential clustering of large dissimilarity data
publishDate 2017
url https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84973519329&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/41653
_version_ 1681422041693028352