Summary
International Symposium on Nonlinear Theory and its Applications
2009
Session Number:C1L-B
Session:
Number:C1L-B2
p-Norm Based Fuzzy c-Means
Takatomi Miyata, Yuchi Kanzawa,
pp.-
Publication Date:2009/10/18
Online ISSN:2188-5079
DOI:10.34385/proc.43.C1L-B2
PDF download (196.3KB)
Summary:
In this paper, we propose two variants of Fuzzy c-Means (FCM) clustering algorithms. Both models are extension of conventional FCM in which not only squared Euclidian norm, L1 norm but also p-norm to the qth power are used as the dissimilarity. One is constructed by using Newton method, and the other, only for the case of p = 2, is constructed by using alternative optimization.