Evolutionary design of artificial neural networks with different nodes
- 23 December 2002
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 670-675
- https://doi.org/10.1109/icec.1996.542681
Abstract
Evolutionary design of artificial neural networks(ANNs) offers a very promising and automatic alternativeto designing ANNs manually. The advantage of evolutionarydesign over the manual design is their adaptability toa dynamic environment. Most research in evolving ANNsonly deals with the topological structure of ANNs and littlehas been done on the evolution of both topological structuresand node transfer functions. This paper presents anew automatic method to design general neural...Keywords
This publication has 10 references indexed in Scilit:
- Combinations of genetic algorithms and neural networks: a survey of the state of the artPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Regularization Theory and Neural Networks ArchitecturesNeural Computation, 1995
- A hybrid algorithm for finding the global minimum of error function of neural networks and its applicationsNeural Networks, 1994
- Distributed and local neural classifiers for phoneme recognitionPattern Recognition Letters, 1994
- An empirical study of genetic operators in genetic algorithmsMicroprocessing and Microprogramming, 1993
- Extrapolation and interpolation in neural network classifiersIEEE Control Systems, 1992
- Approximation by superpositions of a sigmoidal functionMathematics of Control, Signals, and Systems, 1989
- Fast simulated annealingPhysics Letters A, 1987
- An introduction to computing with neural netsIEEE ASSP Magazine, 1987
- Minimization by Random Search TechniquesMathematics of Operations Research, 1981