Evolution of Artificial Neural Networks

Artificial neural networks are computational models of nervous systems. Natural organisms, however, do not possess only nervous systems but also genetic information stored in the nucleus of their cells (genotype). The nervous system is part of the phenotype which is derived from this genotype through a process called development. The information specified in the genotype determines aspects of the nervous system which are expressed as innate behavioral tendencies and predispositions to learn. When neural networks are viewed in the broader biological context of Artificial Life they tend to be accompanied by genotypes and to become members of evolving populations of networks in which genotypes are inherited from parents to offspring (Parisi, 1997). Artificial neural networks can be evolved by using evolutionary algorithms (Holland, 1975; Schwefel, 1995; Koza, 1992). An initial population of different artificial genotype, each encoding the free parameters (e.g. the connection strengths and/or the architecture of the network and/or the learning rules) of a corresponding neural network, are created randomly. The population of networks is evaluated in order to determine the performance (fitness) of each individual network. The fittest networks are allowed to reproduce (sexually or a-sexually) by generating copies of their genotypes with the addition of changes introduced by some genetic operators (e.g., mutations, crossover, duplication). This process is repeated for a number of generations until a network that satisfies the performance criterion (fitness function) set by the experimenter is obtained (for a review of methodological issue see Yao, 1993). The genotype might encode all the free parameters of the corresponding neural network or only the initial value of the parameters and/or other parameters that affects learning. In the former case of the network is entirely innate and there is no learning. In the latter networks changes both philogenetically (i.e. through out generations) and ontogenetically (i.e. during the period of time in which they are evaluated).

[1]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[2]  S. Gould,et al.  Ontogeny and Phylogeny , 1978 .

[3]  John Maynard Smith,et al.  When learning guides evolution , 1987, Nature.

[4]  Peter M. Todd,et al.  Designing Neural Networks using Genetic Algorithms , 1989, ICGA.

[5]  Hiroaki Kitano,et al.  Designing Neural Networks Using Genetic Algorithms with Graph Generation System , 1990, Complex Syst..

[6]  Richard K. Belew,et al.  Evolving networks: using the genetic algorithm with connectionist learning , 1990 .

[7]  David H. Ackley,et al.  Interactions between learning and evolution , 1991 .

[8]  Peter M. Todd,et al.  Exploring adaptive agency II: simulating the evolution of associative learning , 1991 .

[9]  A. Cangelosi,et al.  Cell division and migration in a 'genotype' for neural networks (Cell division and migration in neural networks) , 1993 .

[10]  John R. Koza,et al.  Genetic programming - on the programming of computers by means of natural selection , 1993, Complex adaptive systems.

[11]  Xin Yao,et al.  A review of evolutionary artificial neural networks , 1993, Int. J. Intell. Syst..

[12]  Frank Dellaert,et al.  Toward an evolvable model of development for autonomous agent synthesis , 1994 .

[13]  D. Parisi,et al.  Phenotypic plasticity in evolving neural networks , 1994, Proceedings of PerAc '94. From Perception to Action.

[14]  Frédéric Gruau,et al.  Automatic Definition of Modular Neural Networks , 1994, Adapt. Behav..

[15]  Dale Purves,et al.  Neural Activity And The Growth Of The Brain , 1994 .

[16]  Jeffrey L. Elman,et al.  Learning and Evolution in Neural Networks , 1994, Adapt. Behav..

[17]  Inman Harvey,et al.  The use of genetic algorithms for the development of sensorimotor control systems , 1994, Proceedings of PerAc '94. From Perception to Action.

[18]  Hans-Paul Schwefel,et al.  Evolution and optimum seeking , 1995, Sixth-generation computer technology series.

[19]  Stefano Nolfi,et al.  Learning to Adapt to Changing Environments in Evolving Neural Networks , 1996, Adapt. Behav..

[20]  Giles Mayley,et al.  Landscapes, Learning Costs, and Genetic Assimilation , 1996, Evolutionary Computation.

[21]  D. Parisi.  Artificial Life and Higher Level Cognition , 1997, Brain and Cognition.

[22]  Phil Husbands,et al.  Better Living Through Chemistry: Evolving GasNets for Robot Control , 1998, Connect. Sci..

[23]  Stefano Nolfi,et al.  Co-evolving predator and prey robots , 1998, Artificial Life.

[24]  Stefano Nolfi.  How Learning and Evolution Interact: The Case of a Learning Task which Differs from the Evolutionary Task , 1999, Adapt. Behav..

[25]  Stefano Nolfi,et al.  Learning and Evolution , 1999, Auton. Robots.

[26]  Jeffrey L. Krichmar,et al.  Evolutionary robotics: The biology, intelligence, and technology of self-organizing machines , 2001, Complex..