Title :
Stability analysis of neural networks using stability conditions of fuzzy systems
Author :
Tanaka, Kazuo ; Sano, Manabu
Author_Institution :
Dept. of Mech. Syst. Eng., Kanazawa Univ., Japan
Abstract :
The authors discuss stability of neural networks using stability conditions of fuzzy systems. A parameter region (PR) representation, which graphically shows the location of fuzzy if-then rules in consequent parameter space, is proposed, using the concepts of edge rule (matrix) and minimum representation. The stability criterion of neural networks is illustrated in terms of the PR representation. Some properties for stability of neural networks are derived from the results on the stability criterion
Keywords :
fuzzy logic; neural nets; stability criteria; edge rule; fuzzy if-then rules; fuzzy systems; minimum representation; neural networks; parameter region representation; stability criterion; Fuzzy sets; Fuzzy systems; Input variables; Mechanical systems; Neural networks; Nonlinear systems; Stability analysis; Stability criteria; Sufficient conditions; Systems engineering and theory;
Conference_Titel :
Fuzzy Systems, 1993., Second IEEE International Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-0614-7
DOI :
10.1109/FUZZY.1993.327425
