Title :
Persistence of equilibria under weight variation of feedback continuous-time neural network
Author :
Ling, Bo ; Salam, Fathi M A
Author_Institution :
Dept. of Electr. Eng., Michigan State Univ., East Lansing, MI, USA
Abstract :
For binary patterns, the authors consider the variation of equilibria of the Hopfield-type feedback continuous-time neural network due to perturbations. They show that the equilibria of the feedback continuous-time neural network and its perturbed network are very close as long as the variation of weights is relatively small. The variation of the equilibria can be estimated given the upper bound of the variation of weights
Keywords :
Hopfield neural nets; continuous time systems; learning (artificial intelligence); pattern recognition; Hopfield-type feedback neural net; binary patterns; equilibria variation; feedback continuous-time neural network; perturbations; perturbed network; upper bound; weight variation; Artificial neural networks; Bifurcation; Circuits; Feedforward neural networks; Hardware; Hopfield neural networks; Neural networks; Neurofeedback; Stability; Upper bound;
Conference_Titel :
Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International Symposium on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-1281-3
DOI :
10.1109/ISCAS.1993.394186
