IJPAM: Volume 70, No. 1 (2011)

A NOVEL FUZZY CLUSTERING APPROACH
INDUCED BY COVERING

Zhanhong Shi
College of Science
Gansu Agricultural University
Lanzhou, 730070, P.R. CHINA


Abstract. According to the similarity of objects and the corresponding criteria, we can classify the objects which studied, this mathematical method is called clustering analysis. However, there don't exist clear border between objects usually. In fact, the relationship between them is the fuzzy relationship, we can classify these objects using fuzzy mathematics method, this mathematical method is called fuzzy clustering analysis. In the covering rough set theory, classify the objects of universe of discourse according to the different standard, different coverings of universe can be obtained. In this paper, the problem of how to classify the objects of universe according to these different coverings is discussed. Firstly, the significant degree of covering and the covering degree between the objects in the universe are defined, and by viewing the significant degree of covering as its weight, the corresponding fuzzy matrix is constructed. Secondly, the fuzzy clustering approach induced by covering is built. Finally, a concrete example is employed to illustrate the effectivity of approach which presented in this paper.

Received: January 5, 2011

AMS Subject Classification: 62H30

Key Words and Phrases: fuzzy clustering, covering, significant degree

Download paper from here.



Source: International Journal of Pure and Applied Mathematics
ISSN printed version: 1311-8080
ISSN on-line version: 1314-3395
Year: 2011
Volume: 70
Issue: 1