An Attribute Clustering Algorithm Based on Minimum Spanning Tree
Peng Chun · Systems Engineering - Theory & Practice · 2001
Combining with the method of minimum spanning tree in Graph theory, we present two concepts of similarity and contact degree in this paper, and build a new attribute clustering algorithm based on them. We compare the approach with some other methods such as FCM and AKM, in several problems for analyzing it's effectiveness. It is obvious that the approach we introduced here makes up a loss of other methods and could pactically solve problems in a certain extent.