Abstract
The Freidlin-Wentzell theory deals with the study of random perturbations of dynamical systems. We build several models of genetic algorithms by randomly perturbing simple processes. The asymptotic dynamics of the resulting processes is analyzed with the powerful tools developed by Freidlin and Wentzell and later by Azencott, Catoni and Trouvé in the framework of the generalized simulated annealing. First, a markovian model inspired by Holland's simple genetic algorithm is built by randomly perturbing a very simple selection scheme. The convergence toward the global maxima of the fitness function becomes possible when the population size is greater than a critical value which depends strongly on the optimization problem. In the bitstring case, the critical value of the population size is smaller than a linear function of the chromosome length, provided that the difficulty of the fitness landscape remains bounded. We then use the concepts introduced by Catoni and further generalized by Trouvé to fathom more deeply the dynamics of the two operators mutation-selection algorithm when the population size becomes very large. A new genetic algorithm is finally presented. A new selection mechanism is used, which has the decisive advantage of preserving the diversity of the individuals in the population. When the population size is greater than a critical value, the delicate asymptotic interaction of the perturbations ensures the convergence toward a set of populations which contain all the global maxima of the fitness function.
Preview
Unable to display preview. Download preview PDF.
References
E.H.L. Aarts-P.J.M. Van Laarhoven, Simulated annealing: theory and applications, D. Reidel, Dordrecht-Tokyo, 1987.
O. Catoni, Large deviations for annealing, PhD Thesis, University Paris XI (1990).
O. Catoni, Rough large deviations estimates for simulated annealing, Application to exponential schedules, The Annals of Probability 20 no. 3 (1992), 1109–1146.
O. Catoni, Sharp large deviations estimates for simulated annealing algorithms, Ann. Inst. Henri Poincaré Probab. Statist. 27 no. 3 (1991), 291–383.
R. Cerf, Asymptotic convergence of genetic algorithms, Preprint, submitted (1993).
R. Cerf, The dynamics of mutation-selection algorithms with large population sizes, Ann. Inst. Henri Poincaré Probab. Statist., 1996, to appear.
R. Cerf, A new genetic algorithm, Preprint, submitted (1993). The postscript files of the above three papers are available by anonymous ftp at ftp://blanche.polytechnique.fr/pub/eark/papers/Raphael.Cerf/xxx.ps.gz or with netscape http://blanche.polytechnique.fr/www.eeaax/eeaax.html.
R. Cerf, Une théorie asymptotique des algorithmes génétiques, Thèse, Université Montpellier II (1994). The postscript file of this thesis is available by anonymous ftp at ftp 129.199.96.12, pub/reports/thesis/atga.ps.Z.
T.E. Davis-J.C. Principe, A simulated annealing like convergence theory for the simple genetic algorithm, Proceedings of the 4th ICGA, San Diego, Morgan Kaufmann, 174–181, 1991.
M.I. Freidlin A.D. Wentzell, Random perturbations of dynamical systems, Springer-Verlag, New York, 1984.
D. Goldberg, Genetic algorithms in search, optimization and machine learning, Addison-Wesley, 1989.
C.R. Hwang S.J. Sheu, Singular perturbed Markov chains and exact behaviours of simulated annealing process, Journal of Theoretical Probability Vol. 5 No. 2 (1992), 223–249.
J.H. Holland, Adaptation in natural and artificial systems, The University of Michigan Press, Ann Arbor, 1975.
A.E. Nix M.D. Vose, Modeling genetic algorithms with Markov chains, Annals of Mathematics and AI, 5(1):79–88, 1991.
G. Rudolph, Convergence analysis of canonical genetic algorithms, IEEE Trans on Neural Networks 5(1), 96–101, 1994.
A. Trouvé, Cycle decompositions and simulated annealing, Preprint (1993).
A. Trouvé, Parallélisation massive du recuit simulé, PhD Thesis, University Paris XI (1993).
A. Trouvé, Rough large deviation estimates for the optimal convergence speed exponent of generalized simulated annealing algorithms, Ann. Inst. Henri Poincaré Probab. Statist., 1996, to appear.
Author information
Authors and Affiliations
Corresponding author
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cerf, R. (1996). An asymptotic theory for genetic algorithms. 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_29
Download citation
DOI: https://doi.org/10.1007/3-540-61108-8_29
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