A rapid fuzzy rule clustering method based on granular computing

Abstract

Traditionally, clustering is the task of dividing samples into homogeneous clusters based on their degrees of similarity. As samples are assigned to clusters, users need to manually give descriptions for all clusters. In this paper, a rapid fuzzy rule clustering method based on granular computing is proposed to give descriptions for all clusters. A new and simple unsupervised feature selection method is employed to endow every sample with a suitable description. Exemplar descriptions are selected from sample's descriptions by relative frequency, and data granulation is guided by the selected exemplar fuzzy descriptions. Every cluster is depicted by a single fuzzy rule, which make the clusters understandable for humans. The experimental results show that our proposed model is able to discover fuzzy IF–THEN rules to obtain the potential clusters.