Silhouettes: a graphical aid to the interpretation and validation of cluster analysis
Journal of Computational and Applied Mathematics
Clustering validity checking methods: part II
ACM SIGMOD Record
Connectionist Structures of Type 2 Fuzzy Inference Systems
PPAM '01 Proceedings of the th International Conference on Parallel Processing and Applied Mathematics-Revised Papers
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
New indices for cluster validity assessment
Pattern Recognition Letters
Neuro-fuzzy systems with relation matrix
ICAISC'10 Proceedings of the 10th international conference on Artificial intelligence and soft computing: Part I
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
This paper describes a new cluster validity index for the well-separable clusters in data sets. The validity indices are necessary for many clustering algorithms to assign the naturally existing clusters correctly. In the presented method, to determine the optimal number of clusters in data sets, the new cluster validity index has been used. It has been applied to the complete link hierarchical clustering algorithm. The basis to define the new cluster validity index is founding of the large increments of intercluster and intracluster distances, when the clustering algorithm is performed. The maximum value of the index determines the optimal number of clusters in the given set simultaneously. Obtained results confirm very good performances of the proposed approach.