Towards interval-valued fuzzy set-based collaborative fuzzy clustering algorithms. (September 2018)

• Record Type: Journal Article
• Authors: Ngo, Long Thanh; Dang, Trong Hop; Pedrycz, Witold
• Is Part Of: Pattern recognition. Volume 81(2018:Sep.)
• Page Start: 404
• Page End: 416
• Publication Date: 2018-09
• Subjects: Fuzzy C-Means (FCM) -- Collaborative fuzzy clustering -- Interval - valued fuzzy clustering -- Interval type-2 fuzzy sets -- Clustering validity index; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2018.04.006
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: Propose the interval valued based fuzzy collaborative clustering. Propose the interval valued based fuzzy collaborative clustering with the different number of clusters. Extend the validity index for fuzzy collaborative clustering with interval valued sets. Experiments on various datasets and satellite image segmentation. Abstract: Several studies were devoted to the usage of the Fuzzy C-Means (FCM) algorithm to collaborative clustering, especially in the realm of data analysis, data mining, and pattern recognition. In this study, a novel interval-valued fuzzy set-based approach to realize collaborative clustering is presented. In collaborative clustering diagram, the local clustering results acquired locally (at a specific data site) impact clustering carried out at some other data sites. Those clustering methods endowed with interval-valued fuzzy sets help cope with uncertainties present in the data and the nature of the collaborative process itself. The validity indices such as fuzzy silhouette and SSE (Sum of Squared Error) are extended to quantify results produced by collaborative fuzzy clustering. Several experimental studies are presented using which we demonstrate the advantages of the proposed algorithms.