This Digital Library is published by the Association for Computing. Copyright ¿ 1999 ACM, Inc.

The basic ideas in neural networks

David E. Rumelhart
Bernard Widrow
Michael A. Lehr

Communications of the ACM
Vol. 37, No. 3 (March 1994), Pages 87-92

[Index Terms] ..... [Review]
[Full Text in PDF Format, 2430 KB]


General Terms

DESIGN, THEORY

Categories and Subject Descriptors

F.1.1 Theory of Computation, COMPUTATION BY ABSTRACT DEVICES, Models of Computation, Self-modifying machines.
I.2.6 Computing Methodologies, ARTIFICIAL INTELLIGENCE, Learning, Connectionism and neural nets.
C.1.3 Computer Systems Organization, PROCESSOR ARCHITECTURES, Other Architecture Styles, Neural nets.
A.1 General Literature, INTRODUCTORY AND SURVEY.


From Computing Reviews


J. Tepandi

A long list of scientists motivated to work in the neural network area begins this introductory paper. The paper proceeds with an overview of a basic model of a neuron, a sketch of a brain-like computational device, and a historical overview of neural network work. Learning by example, the backpropagation learning procedure, and generalization are treated in some detail. The paper concludes with hints for successful applications. The references cover a wider than usual span of time, presenting some important earlier work. The paper can be good reading for those wishing to get an overview of basic neural network ideas and some implementation insight in the shortest time with the smallest effort.