A Modified Fuzzy C-Mean Algorithm for Automatic Clustering Number

Abstract:

FCM(Fuzzy C-Means) algorithm is an important algorithm in cluster analysis. It plays an significant role in theory and practice. However, the clustering number of FCM algorithm needs to be set beforehand. This paper proposes an automatic clustering number determination for the classical FCM(Fuzzy C-Means) algorithm. The proposed automatic clustering number determination is based on the cardinality of clustering fuzzy membership used in the CA(Competitive Agglomeration) algorithm. The effectiveness of the proposed algorithm, along with a comparison with CA algorithm, has been showed both qualitatively and quantitatively on a set of real-life datasets.

• Periodical: Applied Mechanics and Materials (Volumes 333-335)
• Pages: 1418-1421
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.333-335.1418
• Online since: July 2013
• Authors: Cheng Quan Huang
• Keywords: Cardinality, Clustering Number, Fuzzy Clustering, Membership

References

[1] RUSPINI E H. Information and Control, 15(1): 22-32(1969).

[2] DUNN J C. Journal of Cybernetics, 3(3): 32-57. (1973).

[3] BEZDEK J C. Pattern Recognition with Fuzzy Objective Function Algorithms[M]. New York: Plenum, (1981).

[4] XU R, WUNSCH D. IEEE Trans on Neural Networks, 16(3): 645-678(2005).

[5] RIGUI H, KRISHNAPURAM R. Pattern Recognition, 30(7): 1109-1119 (1997).

[6] STREHL A, GHOSH J. Journal of Machine Learning Research, 3(12): 583-617 (2002).

[7] RANDA W M. Journal of the American Statistical Association, 66(336): 846-850 (1971).