Miller, G.F., Todd, P.M., and Hegde, S.U. (1989). Designing neural networks using genetic algorithms. In J.D. Schaffer (Ed.), Proceedings of the Third International Conference on Genetic Algorithms (pp. 379-384). San Mateo, CA: Morgan Kaufmann.

Abstract

We present a genetic algorithm method that evolves neural network architectures for specific tasks. Each network architecture is represented as a connection constraint matrix mapped directly into a bit string genotype. Modified standard genetic operators are used during evolution. Architecture fitness is assessed by training particular network instantiations and recording their final performance error. Three applications of this method to simple network mapping tasks are discussed, and we conclude with an indication of possible extensions to this work.



   Back to main page.
   View publications categorized by topic.
   View publication abstracts searchable by keywords