A real-time implementable neural network

Author

Ngolediage, J.E. ; Naguib, R.N.G. ; Dlay, S.S.

Author_Institution

Dept. of Electr. & Electron. Eng., Newcastle upon Tyne Univ., UK

Volume

4

fYear

1994

fDate

27 Jun-2 Jul 1994

Firstpage

2122

Abstract

This paper describes a real-time implementable algorithm that takes advantage of the Lyapunov function, which guarantees an asymptotic behaviour of the solutions to differential equations. The algorithm is designed for feedforward neural networks. Unlike conventional backpropagation, it does not require the suite of derivatives to be propagated from the top layer to the bottom one. Consequently, the amount of circuitry required for an analogue CMOS implementation is minimal. In addition, each unit in the network has its output fed back to itself across a delay element. Results from an HSPICE simulation of the 2.4 micron CMOS architecture are presented

Keywords

CMOS analogue integrated circuits; Lyapunov methods; SPICE; analogue processing circuits; differential equations; feedforward neural nets; neural chips; neural net architecture; real-time systems; 2.4 micron; 2.4 micron CMOS architecture; HSPICE simulation; Lyapunov function; analogue CMOS implementation; asymptotic behaviour; differential equations; feedforward neural networks; output feedback; real-time implementable neural network; Adaptive algorithm; Backpropagation algorithms; Circuits; Equations; Feedforward neural networks; Information processing; Neural networks; Neurons; Robustness; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks, 1994. IEEE World Congress on Computational Intelligence., 1994 IEEE International Conference on

Conference_Location

Orlando, FL

Print_ISBN

0-7803-1901-X

Type

conf

DOI

10.1109/ICNN.1994.374543

Filename

374543