Multi-central general fuzzy clustering model

This paper presents a new model for general fuzzy type-2 clustering called Multi Central General Fuzzy Type-2 Clustering that is an extension of possibilistic c-means (PCM). We mainly focus on uncertainty associated with the cluster centers and define a set of points as the center for each cluster. In our model the degree of belonging is defined as general type-2 fuzzy and there is not any type reduction or defuzzification in the new clustering algorithm. Numerical examples demonstrate the proposed model performance.