Abstract
Evolution Strategies(ES) are an approach to numerical optimization that shows good optimization performance. However, according to our computer simulations, ES shows different optimization performance when a different lower bound of strategy parameters is adopted. We analyze that this is caused by the premature convergence of strategy parameters, although they are traditionally treated as “self-adaptive” parameters. This paper proposes a new extended ES, called RES in order to overcome this brittle property. RES has redundant neutral strategy parameters and adopts new mutation mechanisms in order to utilize the effect of genetic drift to improve the adaptability of strategy parameters. Computer simulations of the proposed approach are conducted using several test functions.
The authors acknowledge financial support through the “Methodology of Emergent Synthesis” project(96P00702) by JSPS (the Japan Society for the Promotion of Science).
This is a preview of subscription content, log in via an institution to check access.
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
T. Bäck (1996), Evolutionary Algorithms in Theory and Practice, Oxford University Press
T. Bäck and H.-P. Schwefel (1993), “An Overview of Evolutionary Algorithms for Parameter Optimization”, Evolutionary Computation, Vol. 1,No.1, pp.1–24
D. Fogel (1995), Evolutionary Computation Toward a New Philosophy of Machine Intelligence, IEEE Press
T. Gohtoh, K. Ohkura and K. Ueda (1996), “An Application of Genetic Algorithm with Neutral Mutations to Job-Shop Scheduling Problems”, Proc. of International Conference on Advances in Production Management Systems, pp. 563–568
D. Goldberg (1989), Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley
K.-H. Liang, X. Yao, Y. Liu, C. Newton and D. Hoffman (1998), “An Experimental Investigation of Self-adaptation in Evolutionary Programming,” Proc. of the Seventh Annual Conference on Evolutionary Programming, Lecture Notes in Computer Science, Springer-Verlag, Berlin, pp.291–300
C. Kappler (1996), “Are Evolutionary Algorithms Improved by Large Mutations?”, H.-M. Voigt, W. Ebeling, I. Rechenberg and H.-P. Schwefel, eds., Proc. Parallel Problem Solving from Nature IV, Vol.1141 of Lecture Notes in Computer Science, Springer-Verlag, pp.346–355
M. Kimura (1983), The Neutral Theory of Molecular Evolution, Cambridge University Press
K. Ohkura and K. Ueda (1995), “Solving Deceptive Problems using Genetic Algorithms with Neutral Mutations”, C. H. Dagli, et al. edited, Intelligent Engineering Systems Through Artificial Neural Networks, Vol.5, ASME Press, pp.345–350
K. Ohkura and K. Ueda (1997), “An Extended Framework for Overcoming Pre-mature Convergence”, Proc. the Seventh International Conference on Genetic Algorithms, Morgan Kaufmann, pp.260–267
X. Yao and Y. Liu (1997), “Fast Evolution Strategies”, Control and Cybernetics, 26(3) pp.467–496
X. Yao and Y. Liu (1997), “Fast Evolution Strategies,”, Proc. of the Sixth Annual Conference on Evolutionary Programming, Lecture Notes in Computer Science, Vol. 1213, Springer-Verlag, Berlin, pp.151–161.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Ohkura, K., Matsumura, Y., Ueda, K. (1999). Robust Evolution Strategies. In: McKay, B., Yao, X., Newton, C.S., Kim, JH., Furuhashi, T. (eds) Simulated Evolution and Learning. SEAL 1998. Lecture Notes in Computer Science(), vol 1585. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-48873-1_3
Download citation
DOI: https://doi.org/10.1007/3-540-48873-1_3
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65907-5
Online ISBN: 978-3-540-48873-6
eBook Packages: Springer Book Archive