Robust Recurrent Credibilistic Modification of the Gustafson - Kessel Algorithm

The task of fuzzy clustering data is very interesting and important problem and often found in many applications related to data mining and exploratory data analysis. For solving these problems the traditional methods require that every vector-observations are fed from data could belong to only one cluster. A more natural is situation when a vector-observations with the various possibilities of membership levels can belong more, than one cluster. In this situation more effective are methods of fuzzy clustering that are synthesized for the allowance of the mutual overlapping of the classes, which are formed in the process of analyzing the data.