k-EVCLUS: Clustering Large Dissimilarity Data in the Belief Function Framework - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

k-EVCLUS: Clustering Large Dissimilarity Data in the Belief Function Framework

Résumé

In evidential clustering, the membership of objects to clusters is considered to be uncertain and is represented by 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 propose to replace the gradient-based optimization procedure in the original EVCLUS algorithm by a much faster iterative row-wise quadratic programming method. We also show that EVCLUS can be provided with only a random sample of the dissimilarities, reducing the time and space complexity from quadratic to linear. These improvements make EVCLUS suitable to cluster large dissimilarity datasets.
Fichier principal
Vignette du fichier
belief2016_paper12.pdf (278.86 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01397949 , version 1 (22-05-2017)

Identifiants

Citer

Orakanya Kanjanatarakul, Songsak Sriboonchitta, Thierry Denoeux. k-EVCLUS: Clustering Large Dissimilarity Data in the Belief Function Framework. 4th International Conference on Belief Functions: Theory and Applications (BELIEF 2016), Sep 2016, Prague, Czech Republic. pp.105-112, ⟨10.1007/978-3-319-45559-4_11⟩. ⟨hal-01397949⟩
136 Consultations
115 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More