Title :
Estimate of the number of equilibria in continuous-time Hopfield neural networks
Author_Institution :
Intelligence Res. Center, Beijing Univ. of Posts & Telecommun., China
Abstract :
A new method is developed to estimate the number of equilibria in continuous-time Hopfield (1982, 1984) neural dynamic systems. By this new method, the number of equilibria in (S) is described clearly when the connection matrix T is upper trigonal, namely, Tij=0 (i>j). Furthermore, it is reasonably conjectured that if T is an arbitrary real matrix, the total number of equilibria in (GS) is no greater than 3a and the total number of asymptotically stable equilibria in (GS) is no greater than 2"
Keywords :
Hopfield neural nets; continuous time systems; matrix algebra; parameter estimation; arbitrary real matrix; asymptotically stable equilibria; continuous-time Hopfield neural dynamic systems; continuous-time Hopfield neural networks; equilibria number estimation; upper triagonal connection matrix; Differential equations; Hopfield neural networks; Intelligent networks; Stability analysis;
Conference_Titel :
Signal Processing Proceedings, 2000. WCCC-ICSP 2000. 5th International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-5747-7
DOI :
10.1109/ICOSP.2000.893408
