Incremental Hierarchical Fuzzy Model Generated from Multilevel Fuzzy Support Vector Regression Network

摘要: Fuzzy rule-based systems are nowadays one of the most successful applications fuzzy logic, but in complex with a large set variables, number rules increases exponentially and the obtained system is scarcely interpretable. Hierarchical alternatives presented in literature to overcome this problem. This paper presents multilevel support vector regression network (MFSVRN) model that learns incremental hierarchical structure based on the Takagi-Sugeno-Kang(TSK) aim coping curse dimensionality and generalization ability. From input–output data pairs, TS-type its parameters are learned by combination clustering linear SVR paper. In addition, an efficient input variable selection method proposed FCM clustering association rules. To achieve high generalization ability, consequence parameters rule learned through new TS-kernel. demonstrates the capabilities MFSVRN conducting simulations function approximations chaotic time-series prediction. also compares simulation results from single-level counterparts- FSVRN Jang's ANFIS model.