Fuzzy clustering algorithms with unknown number of clusters

In the paper the problem of number of clusters in fuzzy clustering algorithms is considered. A novel method for its determination is suggested. The method is based on modification of differential grouping algorithm for determination of centers of fuzzy clusters. The experimental investigations of the suggested algorithm were carried out in the problem UNO counties clustering problem on sustainable development indices. Keywords– fuzzy clustering, fuzzy neural network, index GINI

[1]  Mikhail Z. Zgurovsky,et al.  The Fundamentals of Computational Intelligence: System Approach , 2016, Studies in Computational Intelligence.

[2]  Dong-Chul Park,et al.  Gradient based fuzzy c-means (GBFCM) algorithm , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[3]  R. Yager,et al.  Approximate Clustering Via the Mountain Method , 1994, IEEE Trans. Syst. Man Cybern. Syst..