Abstract
We present PANIC (Parallelism And Neural networks In Classifier systems), a parallel learning system which uses a genetic algorithm to evolve behavioral strategies codified by sets of rules. The fitness of an individual is evaluated through a learning mechanism, QCA (Q-Credit Assignment), to assign credit to rules. QCA evaluates a rule depending on the context where it is applied. This new mechanism, based on Q-learning and implemented through a multi-layer feed-forward neural network, has been devised to solve the rule sharing problem posed by traditional credit assignment methods. To overcome the heavy computational cost of this approach, we propose a decentralized and asynchronous parallel model of the genetic algorithm.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
P. J. Angeline, G. M. Saunders, and J. B. Pollack. An evolutionary algorithm that constructs recurrent neural networks. IEEE Transaction on Neural Networks, 1994.
R. K. Belew, J. Mclnerney, and N. N. Schraudolph. Evolving networks: using the genetic algorithm with connectionist learning. Technical Report CSE CS90-174, Comp. Sci. and Engr., Univ. California, 1990.
D.J. Chalmers. The evolution of learning: An experiment in genetic connectionism. In Proceedings of the 1990 Connectionist Models Summer School, 1990.
A. Giani. Un nuovo approccio alia definizione e all’implement azione di sistemi a classificatori. Master’s thesis, Dip. di Informatica, University of Pisa, Italy, 1992.
A. Giani, F. Baiardi, and A. Starita. Panic: A parallel evolutionary rule-based system. In Proceedings of the Fourth Annual Conference on Evolutionary Programming, 1995.
J. J. Grefenstette. Credit assignment in rule discovery systems based on genetic algorithms. Machine Learning, 3 (23), 1988.
J. J. Grefenstette. A system for learning control strategies with genetic algorithms. In Proceedings of the Third International Conference on Genetic Algorithms and Their Applications, 1989.
J. H. Holland. Escaping brittleness: The possibilities of general-purpose learning algorithm applied to parallel rule-based systems, volume 2 of Machine learning: An artificial inteligence approach. Morgan Kaufmann, 1986.
J. Koza. Genetic programming: On the programming of computers by the means of natural selection. MIT Press, 1992.
L. Lin. Self-improving reactive agents: Case studies of reinforcement learning frameworks. In From Animals to Animats: Proceedings of the First International Conference on Simulation of Adaptive Behaviour, 1990.
R. L. Riolo. Empirical studies of default hierarchies and sequences of rules in learning classifier systems. PhD thesis, University of Michigan, 1988.
D. E. Rumelhart, G. E. Hinton, and R. J. Williams. Learning internal representation by error propagation, volume 1 of Parallel Distributed Processing. MIT Press, 1986.
S. F. Smith. A learning system based on genetic adaptive algorithms. PhD thesis, University of Pittsburgh, 1980.
R. S. Sutton. Learning to predict by the methods of temporal differences. Machine Learning, 3, 1988.
R. S. Sutton. Reinforcement learning architectures for animats. In From Animals to Animats: Proceedings of the First International Conference on Simulation of Adaptive Behaviour, 1990.
G. J. Tesauro. Practical issues in temporal difference learning. Technical Report RC 17223, 1MB T. J. Watson Research Center, Yorktown Heights, NY, 1991.
C. J. C. H. Watkins. Learning with delayed rewards. PhD thesis, University of Cambridge, England, 1989.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1995 Springer-Verlag/Wien
About this paper
Cite this paper
Giani, A., Baiardi, F., Starita, A. (1995). Q-Learning and Parallelism in Evolutionary Rule Based Systems. In: Artificial Neural Nets and Genetic Algorithms. Springer, Vienna. https://doi.org/10.1007/978-3-7091-7535-4_98
Download citation
DOI: https://doi.org/10.1007/978-3-7091-7535-4_98
Publisher Name: Springer, Vienna
Print ISBN: 978-3-211-82692-8
Online ISBN: 978-3-7091-7535-4
eBook Packages: Springer Book Archive