Algorithm for Fuzzy Clustering of Mixed Data with Numeric and Categorical Attributes

Abstract

In many applications numeric as well as categorical features describe the data objects. A variety of algorithms have been proposed for clustering if fuzzy partitions and descriptive cluster prototypes are desired. However, most of these methods are designed for data sets with variables measured in the same scale type (only categorical, or only numeric). We have developed probabilistic distance measure to compute significance of attributes for numeric data, and distance between two categorical values. We used this distance measure with the cluster center definition proposed by Yasser El-Sonbaty and M. A. Ismail [26] to propose Fuzzy-c mean type clustering algorithm for mixed attributes data. The results of the application of the new algorithm show that new technique is quite encouraging.