The Community for Technology Leaders
2013 IEEE Symposium on Computers and Communications (ISCC) (2006)
Cagliari, Sardinia, Italy
June 26, 2006 to June 29, 2006
ISSN: 1530-1346
ISBN: 0-7695-2588-1
pp: 442-447
Mohammad Hossien Yaghmaee , Ferdowsi University of Mashhad, Iran
Ebrahim Bagheri , Ferdowsi University of Mashhad, Iran
Faezeh Ensan , Ferdowsi University of Mashhad, Iran
Clustering belongs to the set of mathematical problems which aim at classification of data or objects into related sets or classes. Many different pattern clustering approaches based on the pattern membership model could be used to classify objects within various classes. Different models of Crisp, Hierarchical, Overlapping and Fuzzy clustering algorithms have been developed which serve different purposes. The main deficiency that most of the algorithms face is that the number of clusters for reaching the optimal arrangement is not automatically calculated and needs user intervention. In this paper we propose a fuzzy clustering technique (FACT) which determines the number of appropriate clusters based on the pattern essence. Different experiments for algorithm evaluation were performed which show a much better performance compared with the typical widely used K-means clustering algorithm
Mohammad Hossien Yaghmaee, Ebrahim Bagheri, Faezeh Ensan, "FACT: A New Fuzzy Adaptive Clustering Technique", 2013 IEEE Symposium on Computers and Communications (ISCC), vol. 00, no. , pp. 442-447, 2006, doi:10.1109/ISCC.2006.73
105 ms
(Ver 3.3 (11022016))