An integrated method for the construction of compact fuzzy neural models

Authors:
Wanqing Zhao;Kang Li;George W. Irwin;Minrui Fei
Affiliations:
Intelligent Systems and Control, School of Electronics, Electrical Engineering and Computer Science, Queen's University Belfast, Belfast, UK;Intelligent Systems and Control, School of Electronics, Electrical Engineering and Computer Science, Queen's University Belfast, Belfast, UK;Intelligent Systems and Control, School of Electronics, Electrical Engineering and Computer Science, Queen's University Belfast, Belfast, UK;Shanghai Key Laboratory of Power Station Automation Technology, School of Mechatronical Engineering and Automation, Shanghai University, Shanghai, China
Venue:
ICIC'10 Proceedings of the 6th international conference on Advanced intelligent computing theories and applications: intelligent computing
Year:
2010

Citing 4
Cited 0

Training ANFIS as an identifier with intelligent hybrid stable learning algorithm based on particle swarm optimization and extended Kalman filter

Fuzzy Sets and Systems
Orthogonal-least-squares regression: A unified approach for data modelling

Neurocomputing
Robust neurofuzzy rule base knowledge extraction and estimation using subspace decomposition combined with regularization and D-optimality

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
DENFIS: dynamic evolving neural-fuzzy inference system and its application for time-series prediction

IEEE Transactions on Fuzzy Systems

Quantified Score

Hi-index	0.02

Visualization

Abstract

To construct a compact fuzzy neural model with an appropriate number of inputs and rules is still a challenging problem. To reduce the number of basis vectors most existing methods select significant terms from the rule consequents, regardless of the structure and parameters in the premise. In this paper, a new integrated method for structure selection and parameter learning algorithm is proposed. The selection takes into account both the premise and consequent structures, thereby achieving simultaneously a more effective reduction in local model inputs relating to each rule, the total number of fuzzy rules, and the whole network inputs. Simulation results are presented which confirm the efficacy and superiority of the proposed method over some existing approaches.