On Entropy Based Fuzzy Non-Metric Model with A Variable Controlling Cluster Sizes

Abstract

The fuzzy non-metric model is one of the clustering methods in which the membership grade of each datum to each cluster is calculated directly from dissimilarities between data. In conventional fuzzy c-means clustering, a variables for controlling cluster sizes are introduced for handling a set of data which has different sizes of clusters. We propose entropy based fuzzy non-metric model with a variable controlling cluster sizes and construct its algorithm in this paper.