Fuzzy set-based distant cluster identification
Keon Myung Lee, Sun A. Lee · Annual Conference on Computers · 2010
This presents a method to identify distant clusters for a data set using fuzzy techniques. The proposed method first applies a fuzzy clustering algorithm to the data set. It introduces a fuzzy membership function called distance membership function which transforms metric distance between two objects into the degree of farness. In order to measure the degree of farness between two fuzzy clusters, it defines the cluster farness measures which incorporate distance membership function and fuzzy clusters. It uses the measures to construct a fuzzy distant relation for clusters from which the distant clusters are identified.