Derivatives of Fuzzy C-means method and their application comparisons

Author

Yan, Chunjuan

Author_Institution

Fac. of Inf., Qingdao Univ. of Sci. & Technol., Qingdao, China

fYear

2011

fDate

27-29 June 2011

Firstpage

326

Lastpage

329

Abstract

Fuzzy C-means (FCM) is also called soft K-means, which is a wildly used unsupervised clustering method. Its derivatives comes out for different requirements, in this paper we compare four related clustering algorithms, which includes FCM and traditional Relational Fuzzy C-means (RFCM) and None Euclidean Relational Fuzzy C-means (NERFCM) and Any Relational Fuzzy C-means (ARFCM). Their common points and different limitations on usage are discussed, finally an optimal clustering algorithm is chosen for application on human posture classification, and experiments prove its efficiency and sensitivity.

Keywords

fuzzy set theory; pattern clustering; unsupervised learning; ARFCM; NERFCM; RFCM; any relational fuzzy C-means; human posture classification; none Euclidean relational fuzzy C-means; optimal clustering algorithm; relational fuzzy C-means; soft K-means; unsupervised clustering method; Algorithm design and analysis; Classification algorithms; Clustering algorithms; Convergence; Histograms; Humans; Prototypes; ARFCM; NERF C-means; fuzzy C-mean;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Service System (CSSS), 2011 International Conference on

Conference_Location

Nanjing

Print_ISBN

978-1-4244-9762-1

Type

conf

DOI

10.1109/CSSS.2011.5972174

Filename

5972174