A Rough Set-Based Hierarchical Clustering Algorithm for Categorical Data
Duo Chen, Cui Du-wu, Chaoxue Wang, Zhurong Wang · 2006
In this paper, rough set theory is applied to the clustering analysis. The clustering decision table is formed through the introduction of decision attribute into data table, thereby further defining the attribute membership matrix. The consistent degree and aggregate degree are present, and their functions in the clustering process are deeply analyzed. The clustering level calculation formula is designed, in which two factors such as consistent degree and aggregate degree are taken into comprehensive account. Also, this paper gives the categorical similarity measure based on Euclidean distance so as to better solve the problem of difficult measurement of categorical data because of the non-numerical data nature. On the basis of the above work, a novel categorical clustering algorithm is designed.