Analysis of a simple hysteresis network and its application for an effective associative memory

The authors analyze the stable equilibria and stable periodic outputs for a simple hysteresis network. The simplified network includes a binary hysteresis function and has only three parameters. It is shown that the number of stable equilibria and the stable periodic output sequences can be controlled by the three parameters. The domain of attraction of stable equilibria is completely clarified and the local domain of attraction of periodic outputs is given. An application is proposed for an effective associative memory. This associative memory guarantees storing of all desired memories. It has no spurious stable output and it recalls the closest desired memory from an input vector.<<ETX>>

[1]  Joos Vandewalle,et al.  A new model of neural associative memories , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[2]  T. Saito,et al.  Dynamics of hysteretic neural networks , 1991, 1991., IEEE International Sympoisum on Circuits and Systems.

[3]  Anthony N. Michel,et al.  A synthesis procedure for Hopfield's continuous-time associative memory , 1990 .

[4]  Wolfgang Porod,et al.  Analysis and synthesis of a class of neural networks: variable structure systems with infinite grain , 1989 .

[5]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[6]  Toshimichi Saito,et al.  Chaos and fractals from a forced artificial neural cell , 1993, IEEE Trans. Neural Networks.

[7]  T. Saito,et al.  Analysis and synthesis of a continuous-time hysteresis neural network , 1992, [Proceedings] 1992 IEEE International Symposium on Circuits and Systems.

[8]  Kenya Jin'no,et al.  Analysis and synthesis of continuous‐time hysteretic neural networks , 1993 .

[9]  T. Saito,et al.  On inner connection of artificial neural systems , 1992, [Proceedings] 1992 IEEE International Symposium on Circuits and Systems.