In this paper, we propose an evolutionary approach to the design of optimal modular neural network architectures. In this approach, a modular neural network is treated as a phenotype of an individual, and the modular architecture is optimized through the evolution of its genetic representation (genotype) by using genetic algorithms. As one of the modular neural networks, we adopt Cross-Coupled Hopfield Nets (CCHN) in which plural Hopfield networks are coupled to each other. The architecture of the CCHN is represented by some structural-parameters such as the number of modules, the numbers of module units, and the module connectivity. These parameters for an individual are encoded in a binary string. In the simulation, our genetic system is applied to associative memories. The fitness of an individual is defined so as to be larger when the individual has a simpler architecture as well as when the association performance is higher. In the simulation, we verify that the genetic system finds highperformance individuals with simple modular architectures.
[1]
Bart L. M. Happel,et al.
Design and evolution of modular neural network architectures
,
1994,
Neural Networks.
[2]
J J Hopfield,et al.
Neurons with graded response have collective computational properties like those of two-state neurons.
,
1984,
Proceedings of the National Academy of Sciences of the United States of America.
[3]
Kazuyoshi Tsutsumi,et al.
An Autoassociative Memory Model Derived from a Modular Neural Network and a Diversity of the Association Properties
,
1997
.
[4]
K. Tsutsumi,et al.
Cross-coupled Hopfield nets via generalized-delta-rule-based internetworks
,
1990,
1990 IJCNN International Joint Conference on Neural Networks.
[5]
Kazuyoshi Tsutsumi,et al.
An artificial modular neural network and its basic dynamical characteristics
,
1998,
Biological Cybernetics.