Clustering Based on Fuzzy Rule-Based Classifier
Clustering is the unsupervised classification of patterns which has been addressed in many contexts and by researchers in many disciplines. Fuzzy clustering is recommended than crisp clustering when the boundaries among the clusters are vague and uncertain. Popular clustering algorithms are K-means, K-medoids, Hierarchical Clustering, fuzzy-c-means and their variations. But they are sensitive to number of potential clusters and initial centroids. Fuzzy rule based Classifier is supervised and is not sensitive to number of potential clusters. By taking the advantages of supervised classification, this paper intended to design an unsupervised clustering algorithm using supervised fuzzy rule based classifier. Fuzzy rule with certainty grade plays vital role in optimizing the rule base which is exploited in this paper. The proposed classifier and clustering algorithm have been implemented in Matlab R2010a and tested with various benchmarked multidimensional datasets. Performance of the proposed algorithm is compared with other popular baseline algorithms.
KeywordsClustering Classification Fuzzy clustering Fuzzy rule-based classifier
- 4.Yager, R., Filev, D.: Generation of fuzzy rules by mountain clustering. J. Intel. Fuzzy Syst. 2, 209–211 (1994)Google Scholar
- 6.Yuan, B., Klir, G.J., Stone, J.F.: Evolutionary fuzzy c-means clustering algorithm. In: Fuzzy-IEEE, pp. 2221–2226 (1995)Google Scholar
- 16.Klir, G.J., Yuan, B.: Fuzzy Sets and Fuzzy Logic: Theory and Applications. In Pattern Recognition. Prentice-Hall, Englewood Cliffs (1995)Google Scholar