One of the main drawbacks of the FCM clustering algorithm is that it does not calculate the suitable number of clusters. This paper presents a method to solve this problem, by means of an equalization function (using uniform data) for the FCM functional J. The results for 2 and 3 dimensional data tests are also presented.

External cluster validity indices (CVIs) are used to quantify the quality of a clustering by comparing the similarity between the clustering and a ground truth partition. However, some external CVIs show a biased behaviour when selecting the most similar clustering. Users may consequently be misguided by such results. Recognizing and understanding the bias behaviour of CVIs is therefore crucial...

This talk is about three visual cluster validity methods developed by the authors that can be used for both object and relational data sets. The original method VAT (visual assessment of clustering tendency) works nicely for dissimilarity data up to about n = 5000 objects, but VAT quickly bumps up against storage and resolution limits, and is of limited utility for large data sets. The second m...

the new sample, that is Thus, in general, our learning algorithm may require more than rlog,(IXI-1)1 + 1 samples to solve the problem in the worst case. We have analyzed the problem of constructing a linear classifier for a finite set X of linearly separable vectors by partially supervised leaming. The proposed learning algorithm consists of two major operations: sample selection and classifier...

With the invention of biotechnological high throughput methods like DNA microarrays and the analysis of the resulting huge amounts of biological data, clustering algorithms gain new popularity. In practice the question arises, which clustering algorithm as well as which parameter set generates the most promising results. Little work is addressed to the question of evaluating and comparing the c...

Clustering methods serve as common tools for efficient data analysis in many fields of science. The essential, yet often neglected, step in the cluster analysis is validation of the clustering results. This paper presents a novel cluster validity index, which is the modification of the well-known Dunn’s index. Our proposal is based on its generalization considering the shortest paths between da...

This paper addresses two most important issues in cluster analysis. The first issue pertains to the problem of deciding if two objects can be included in the same cluster. We propose a new similarity decision methodology which involves the idea of cluster validity index. The proposed methodology replaces a qualitative cluster recognition process with a quantitative comparison-based decision pro...

Research conducted under the supervision of Prof. Eytan Domany November1999 1 Acknowledgments I would like to thank Professor Eytan Domany for an intriguing year. His devoted guidance and interesting suggestions had taught m e a l o t. I also wish to thank Gaddy Getz and Noam Shental for useful discussions and ideas. I really enjoyed working with them. To my family and friends I am indebted for...

Clustering is a process of discovering groups of objects such that the objects of the same group are similar, and objects belonging to different groups are dissimilar. A number of clustering algorithms exist that can solve the problem of clustering, but most of them are very sensitive to their input parameters. Therefore it is very important to evaluate the result of them. The minimum spanning ...