Title :
On dynamics of a learning feedback associative memory
Author :
Zeng, Huanglin ; Swiniarski, Roman W.
Author_Institution :
Sichuan Inst. of Light Ind. & Chem. Technol., China
fDate :
30 Apr-3 May 1995
Abstract :
In this paper, we present conditions in which a dynamical feedback associative neural network can be treated as a time-invariant system, and point out confinements on changing rate of a dynamical associative network. Stability constraints to guarantee a specified stable equilibria of an associative neural network with slowly synapse-varying structures are derived. The exponential stability and trajectory bounds of motions of the network equilibria under arbitrary structural perturbations are investigated
Keywords :
asymptotic stability; content-addressable storage; learning (artificial intelligence); recurrent neural nets; dynamics; exponential stability; learning feedback associative memory; neural network; synapse-varying structures; time-invariant system; trajectories; Associative memory; Chemical industry; Chemical technology; Equations; Feedback; Jacobian matrices; Neural networks; Neurofeedback; Neurons; Stability;
Conference_Titel :
Circuits and Systems, 1995. ISCAS '95., 1995 IEEE International Symposium on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-2570-2
DOI :
10.1109/ISCAS.1995.520346
