Learning stable mutation in (1, lambda)ES evolutionary strategy

• Autorzy: Prętki P.
• Konferencja: Evolutionary Computation and Global Optimization (10; Krajowa Konferencja Algorytmy Ewolucyjne i Optymalizacja Globalna; 11-13.06.2007; Będlewo, Poland)
• Języki publikacji: EN

Abstrakty

EN

In this paper the concept of two-dimensional discrete spectral measure is introduced in the context of its application to real-valued evolutionary strategy (1, lambda)ES. The notion of discrete spectral measure makes it possible to uniquely define a class of multivariate heavy-tailed distributions, that have received more and more attention of evolutionary optimization community, recently. In particular, an adaptation procedure known from the class of estimation of distribution algorithms (EDAs) is proposed. The effectiveness of the evolutionary strategy is tested by means of a set popular benchmark functions.

Słowa kluczowe

EN

evolutionary strategy; discrete spectral measure

• Wydawca: Oficyna Wydawnicza Politechniki Warszawskiej
• Czasopismo: Prace Naukowe Politechniki Warszawskiej. Elektronika
• Rocznik: 2007
• Tom: z. 160
• Strony: 233--240
• Opis fizyczny: Bibliogr. 16 poz., wykr.

Twórcy

autor

Prętki P.

• University of Zielona Góra,

Bibliografia

• [1] H.G. Beyer and H.P. Schwefel. Evolution strategies - a comprehensive introduction. Natural Computing, 1(1):3-52, 2002.
• [2] M. Gutowski. Lévy flights as an underlying mechanism for a global optimization algorithm. In 5th Conf. Evolutionary Algorithms and Global Optimization, Jastrzebia Gora, Poland, 2001. Wydział Elektroniki i Technik Informacyjnych Politechniki Warszawskiej, Warsaw University of Technology Press.
• [3] N. Hansen and A. Ostermeier. Completely derandomized self-adaptation in evolution strategies. Evolutionary Computation, 9(2):159-195, 2001.
• [4] S. Kern, S. Uller, D. Uche, N. Hansen, and P. Koumoutsakos. Learning probability distributions in continuous evolutionary algorithms. In Workshop on Fundamentals in Evolutionary Algorithms, Thirtieth International Colloquium on Automata, Languages and Programming, Eindhoven, 2004.
• [5] P. Larranaga and J.A. Lozano. Estimation of Distribution Algorithms: A New Tool for Evolutionary Optimization. Kluwer Academic Publishers, Boston, USA, 2001.
• [6] Z. Michalewicz. Genetic algorithms+data structures=evolution programs (3rd ed.). Springer-Verlag, London, UK, 1996.
• [7] Z. Michalewicz. Evolutionary Algorithms in Engineering Applications. Springer-Verlag New York, Inc., Secaucus, NJ, USA, 1997.
• [8] J.P. Nolan. Stable Distributions, Models for Heavy Tailed Data. Springer-Verlag, Berlin Heidelberg, 2002.
• [9] J.P. Nolan, A.K. Panorska, and J.H. McCulloch. Estimation of stable spectral measures, stable non-gaussian models in finanse and econometrics. Math. Comput. Modelling, 34(9):1113-1122, 2001.
• [10] A. Obuchowicz. Evolutionary Algorithms in Global Optimization and Dynamic System Diagnosis. Lubuskie Scientific Society, Zielona Góra, Zielona Góra, Poland, 2003.
• [11] A. Obuchowicz and P. Pretki. Phenotypic evolution with a mutation based on symmetric α-stable distributions. International Journal of Applied Mathematics and Computer Science, 14(3):289-316, 2004.
• [12] G. Rudolph. Convergence of evolutionary algorithms in general search spaces. In Proc. ICEC'96, Nagoya, pages 50-54, 1996.
• [13] G. Samorodnitsky and M.S. Taqqu. Stable Non-Gaussian Random Processes. Chapman and Hall, New York, 1994.
• [14] X. Yao and Y. Liu. Fast evolutionary programming. In L J Fogel, P J Angeline, and T Bäck, editors, Proc. 5th Ann. Conf. on Evolutionary Programming, Cambridge, MA, 1996. MIT Press.
• [15] X. Yao and Y. Liu. Fast evolution strategies. In Peter J. Angeline, Robert G. Reynolds, John R. McDonnell and Russ Eberhart, editors, Evolutionary Programming VI, pages 151-161, Berlin, 1997. Springer.
• [16] A. Zolotariev. One-Dimensional Stable Distributions. American Mathematical Society, Providence, 1986.