Generating network trajectories using gradient descent in state space

Author

Hahnloser, Richard H R

Author_Institution

Inst. for Theor. Phys., Eidgenossische Tech. Hochschule, Zurich, Switzerland

Volume

3

fYear

1998

fDate

4-9 May 1998

Firstpage

2373

Abstract

A local and simple learning algorithm is introduced that gradually minimizes an error function for neural states of a general network. Unlike standard backpropagation algorithms, it is based on linearizing the neurodynamics which are interpreted as constraints for the different network variables. From the resulting equations, the weight update is deduced which has a minimal norm and produces state changes directed precisely towards target values. As an application, it is shown how to generate desired neural state space curves on recurrent Hopfield-type networks

Keywords

Hopfield neural nets; correlation methods; learning (artificial intelligence); state-space methods; Hopfield-type networks; gradient descent method; learning algorithm; network trajectories; neurodynamics; recurrent neural networks; state space curves; tangential correlation algorithm; weight update; Backpropagation algorithms; Computer architecture; Computer networks; Intelligent networks; Neural networks; Neurodynamics; Neurons; Nonlinear equations; Physics; State-space methods;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on

Conference_Location

Anchorage, AK

ISSN

1098-7576

Print_ISBN

0-7803-4859-1

Type

conf

DOI

10.1109/IJCNN.1998.687233

Filename

687233