This paper presents fuzzy clustering algorithm for fuzzy data based on alpha-cuts. A new suitable definition for distance between two arbitrary fuzzy numbers based on alpha-cuts is proposed. We then reformulate fuzzy c-means (FCM) with fuzzy data and fuzzy centers based on alpha-cuts. The effectiveness of the proposed clustering algorithm is tested for two fuzzy data sets and then it is compared with two other methods; the fuzzy c-number (FCN) algorithm and the mixed-type variables FCM (MVFCM) algorithm.

Keywords