Title :
Training asymptotically stable recurrent neural networks
Author :
Dimopoulos, Nikitas J. ; Neville, Stephen ; Dorocicz, John T. ; Jubien, Chris
Author_Institution :
Dept. of Electr. & Comput. Eng., Victoria Univ., BC, Canada
Abstract :
Presents a class of recurrent networks which are asymptotically stable. For these networks, the authors discuss their similarity with certain structures in the central nervous system, and prove that if an interconnection pattern that does not allow excitatory feedback is used, then the resulting recurrent neural network is stable. The authors introduce a training methodology for networks belonging to this class, and use it to train networks that successfully identify a number nonlinear systems
Keywords :
asymptotic stability; identification; learning (artificial intelligence); nonlinear systems; recurrent neural nets; asymptotically stable recurrent neural networks; central nervous system; interconnection pattern; nonlinear systems; training methodology; Biological neural networks; Central nervous system; Nerve fibers; Nervous system; Neural networks; Neurofeedback; Neurons; Organisms; Recurrent neural networks; Stability;
Conference_Titel :
Systems, Man and Cybernetics, 1995. Intelligent Systems for the 21st Century., IEEE International Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-2559-1
DOI :
10.1109/ICSMC.1995.538485
