Title :
The stability of the analog Hopfield neural network with iterative dynamics
Author_Institution :
Sch. of Comput. & Math., Deakin Univ., Clayton, Vic., Australia
Abstract :
We first present an example which shows that a general dynamic system with an infinite number of fixed points may never stabilize, even when there exists a Lyapunov function which strictly decreases along all possible trajectories. We then explain why the analog Hopfield neural network with iterative dynamics always converges to a fixed point, regardless of how many fixed points it has
Keywords :
Hopfield neural nets; Lyapunov methods; analogue processing circuits; convergence; functions; iterative methods; stability; analog Hopfield neural net; convergence; fixed points; general dynamic system; iterative dynamics; stability; strictly decreasing Lyapunov function; trajectories; Convergence; Counting circuits; Hopfield neural networks; Microwave integrated circuits; Neural networks; Neurons; Stability;
Conference_Titel :
Intelligent Processing Systems, 1997. ICIPS '97. 1997 IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-4253-4
DOI :
10.1109/ICIPS.1997.672821
