Abstract
This paper deals with technical issues relevant to artificial neural net (ANN) training by genetic algorithms. Neural nets have applications ranging from perception to control; in the context of control, achieving great precision is more critical than in pattern recognition or classification tasks. In previous work, the authors have found that when employing genetic search to train a net, both precision and training speed can be greatly enhanced by an input renormalization technique. In this paper we investigate the automatic tuning of such renormalization coefficients, as well as the tuning of the slopes of the transfer functions of the individual neurons in the net. Waiting time analysis is presented as an alternative to the classical ”mean performance” interpretation of GA experiments. It is felt that it provides a more realistic evaluation of the real-world usefulness of a GA.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
P. J. Angeline, G. M. Saunders and J. B. Pollack, An evolutionary algorithm that construct recurrent neural networks. To appear in IEEE Transactions on Neural Networks.
D. E. Goldberg, Genetic algorithms in search, optimization and machine learning, Addison Wesley, 1989.
S. A. Harp and T. Samad, Genetic synthesis of neural network architecture, in Handbook of Genetic Algorithms, L. Davis Ed., Van Nostrand Reinhold, New York, 1991.
R. Hecht-Nielsen, Neurocomputing, Addison Wesley, 1990.
J. Holland, Adaptation in natural and artificial systems, University of Michigan Press, Ann Harbor, 1975.
Empirical studies on the speed of convergence of neural network training using genetic algorithms. In Proceedings of Eight National Conference on Artificial Intelligence, AAAI-90, Vol. 2, pp 789–795, Boston, MA, 29 July–3 Aug 1990. MIT Press, Cambridge, MA.
G. Mani, Learning by Gradient Descent in Function Space, in Proceedings of IEEE Conference on System, Man, and Cybernetics, pp. 242–247, Los Angeles 1990.
Z. Michalewicz, Genetic Algorithms + Data Structures = Evolution Programs, Springer Verlag 1992.
D. Nguyen, B. Widrow, The truck Backer Upper: An example of self learning in neural networks, in Neural networks for Control, W. T. Miller III, R. S. Sutton, P. J. Werbos eds, The MIT Press, Cambridge MA, 1990.
N. J. Radcliffe, Equivalence Class Analysis of Genetic Algorithms, in Complex Systems 5, pp 183–205, 1991.
E. Ronald, M. Schoenauer Genetic lander: An experiment in accurate neuro-genetic control, in PPSN 94, to appear.
D. E. Rumelhart, J. L. McClelland, Parallel Distributed Processing — Exploration in the micro structure of cognition, MIT Press, Cambridge MA, 1986.
J. D. Schaffer, R. A. Caruana and L. J. Eshelman, Using genetic search to exploit the emergent behavior of neural networks, Physica D 42 (1990), pp244–248.
M. Schoenauer, E.Ronald S.Damour, Evolving Networks for Control, in Neuronimes 93, EC2, Paris 1993.
M. Schoenauer, E.Ronald Neuro Genetic Truck Backer-Upper Controller, in IEEE World Conference on Computational Intelligence, Orlando 1994.
D. G. Stork, S. Walker, M. Burns & B. Jackson, Preadaptation in Neural Circuits, in Proceedings of International Joint Conference on Neural Networks 8, pp 1–202–1–205, Erlbaum, Hillsdale 1990.
Steve C. Suddarth, The symbolic-neural method for creating models and control behaviors from examples. Ph.D. dissertation. University of Washington. 1988.
Xin Yao, A Review of Evolutionary Artificial Neural Networks, in International Journal of Intelligent Systems 8, pp 539–567, 1993.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Schoenauer, M., Ronald, E. (1996). How long does it take to evolve a neural net?. In: Alliot, JM., Lutton, E., Ronald, E., Schoenauer, M., Snyers, D. (eds) Artificial Evolution. AE 1995. Lecture Notes in Computer Science, vol 1063. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61108-8_41
Download citation
DOI: https://doi.org/10.1007/3-540-61108-8_41
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61108-0
Online ISBN: 978-3-540-49948-0
eBook Packages: Springer Book Archive