Fuzzy type-2 c-ellipses clustering
S. MalekMohamadi Golsefid, M.H. Fazel Zarandi, Susan Bastani · 2013
In this paper, a new prototype for fuzzy type-2 clustering is proposed. The proposed model considers ellipsoid shape for clusters and ellipsoid focal vectors are basis for measuring the distance from clusters, instead of cluster centers. Therefore, the degrees of belonging to clusters are defined as interval fuzzy type-2 sets indicated by the upper and lower membership values. The results of proposed model on the sample data sets are compared with fuzzy c-ellipsoid clustering algorithms.