Abstract
The chapter claims that the search distributions of Estimation of Distribution Algorithms (EDAs) contain much information that can be obtained with the help of modern statistical techniques to create powerful strategies for expensive optimization. For example, it shows how the regularization of some parameters of the EDAs probabilistic models can yield dramatic improvements in efficiency. In this context a new class, Shrinkage EDAs, based on shrinkage estimation is presented. Also, a novelmutation operator based on a regularization of the entropy is discussed. Another key contribution of the chapter is the development of a new surrogate fitness model based on the search distributions. With this method the evolution starts in the fitness landscape, switches to the log-probability landscape of the model and then backtracks to continue in the original landscape if the optimum is not found. For the sake of completeness the chapter reviews other techniques for improving the sampling efficiency of EDAs. The theoretical presentation is accompanied by numerical simulations that support the main claims of the chapter.
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
Beinlich, I., Chavez, H.R., Cooper, G.: The ALARM Monitoring System: a Case Study with two Probabilistic Inference Techniques for Belief Networks. In: Artificial Intelligence in Medical Care, pp. 247–256 (1989)
Blanco, R., Lozano, J.A.: Empirical comparison of Estimation of Distribution Algorithms in combinatorial optimization. In: Larrañaga, P., Lozano, J.A. (eds.) Estimation of Distribution Algorithms. A New Tool for Evolutionar y Computation, Kluwer Academic Publishers, Dordrecht (2002)
Brown, L.E., Tsamardinos, I., Aliferis, C.F.: A comparison of novel and state-of-the-art polynomial bayesian network learning algorithms. In: AAAI, pp. 739–745 (2005)
Chiba, K., Jeong, S., Shigeru, O., Morino, H.: Data mining for multidisciplinary design space of regional-jet wing. In: Proceedings the 2005 IEEE Congress on Evolutionary Computation CEC 2005, pp. 2333–23340 (2005)
Deb, K., Goldberg, D.E.: Analyzing deception in trap functions. In: Analyzing deception in trap functions. Foundations of Genetic Algorithms, vol. 2, pp. 93–108 (1993)
Eby, D., Averill, R.C., Punch, W.F.I., Goodman, E.D.: Evaluation of injection island ga performance on flywheel design optimization. In: Proceedings of the Third Conference on Adaptive Computing in Design and Manufacturing (1998)
Groppo, T.: Learning bayesian networks skeleton: A comparison between TPDA and PMMS algorithm. Master’s thesis, Universite Claude Bernard Lyon I (2006)
Ireland, C.T., Kullback, S.: Contingency Tables with Given Marginals. Biometrika 55, 179–188 (1968)
Jaynes, E.T.: Information Theory and Statistical Mechanics. Physics Review 6, 620–643 (1957)
Jiroušek, R., Přeučil, S.: On the Effective Implementation of the Iterative Proportional Fitting Procedure. Comput. Statistics and Data Analysis 19, 177–189 (1995)
Kim, H.S., Cho, S.B.: An efficient genetic algorithm with less fitness evaluation by clustering. In: Proceedings of 2001 IEEE Conference on Evolutionary Computation, pp. 887–894 (2001)
Larrañaga, P., Lozano, J.A. (eds.): Estimation of Distribution Algorithms. A New Tool for Evolutionary Computation. Kluwer Academic Publishers, Dordrecht (2002)
Larrañaga, P., Lozano, J.A., Miqueles, T., E-Bengoetxea: Experimental Results in Function Optimization with EDAs in Continuos Domain. In: Larrañaga, P., Lozano, J.A. (eds.) Estimation of Distribution Algorithms. A New Tool for Evolutionary Computation. Kluwer Academic Publishers, Dordrecht (2002)
Lauritzen, S.L.: Graphical Models. Oxford Press (1996)
Ledoit, O., Wolf, M.: Improved estimation of the covariance matrix of stock returns with an application to portfolio selection. J. Empir. Finance 10, 603–621 (2003)
Lewis, P.M.: Approximating Probability Distributions to Reduce Storage Requirements. Information and Control 2, 214–225 (1959)
Liang, K.H., Yao, X., Newton, C.: Evolutionary search of approximated n-dimensional landscapes. International Journal of Knowledge-Based Intelligent Engineering Systems 4(3), 172–183 (2000)
Lozano, J., Larrañaga, P., Inza, I., Bengoetxea, E. (eds.): Towards a new Evolutionary Computation. Advances on Estimation of Distribution Algorithms. Studies in Fuzziness and Soft Computing. Springer, Heidelberg (2006)
Madera, J.: Hacia una Generación Eficiente de Algoritmos Evolutivos con Estimación de Distribuciones: Pruebas de (In)dependencia +Paralelismo. PhD thesis, Instituto de Cibernética, Matemática y Física, La Habana (in Spanish). Adviser: A. Ochoa (2009)
Madera, J., Ochoa, A.: Un EDA basado en aprendizaje Max-Min con escalador de colinas. Tech. Rep. 396, ICIMAF (2006)
Madera, J., Ochoa, A.: Algoritmos Evolutivos con Estimación de Distribuciones Bayesianas basados en pruebas de (in)dependencia. Tech. Rep. 474, ICIMAF (2008)
Mahnig, T., Mühlenbein, H.: Comparing the Adaptive Boltzmann Selection Schedule SDS to Truncation Selection. In: Third International Symposium on Adaptive Systems, ISAS 2001, Evolutionary Computation and Probabilistic Graphical Models, La Habana, pp. 121–128 (2001)
Meyer, C.H.: Korrektes Schlies̈en bei Unvollständiger Information. PhD thesis, Fernuniversität Hagen (1998) (in German)
Mühlenbein, H., Mahnig, T.: Mathematical analysis of evolutionary algorithms for optimization. In: Proceedings of the Third International Symposium on Adaptive Systems, pp. 166–185 (2001)
Mühlenbein, H., Mahnig, T.: Evolutionary Optimization and the Estimation of Search Distributions. Journal of Approximate Reasoning 31(3), 157–192 (2002)
Mühlenbein, H., Paas, G.: From Recombination of Genes to the Estimation of Distributions I. Binary Parameters. In: Ebeling, W., Rechenberg, I., Voigt, H.-M., Schwefel, H.-P. (eds.) PPSN 1996. LNCS, vol. 1141, pp. 178–187. Springer, Heidelberg (1996)
Mühlenbein, H., Mahnig, T., Ochoa, A.: Schemata, Distributions and Graphical Models in Evolutionary Optimization. Journal of Heuristics 5(2), 213–247 (1999)
Nilsson, D.: An Efficient Algorithm for Finding the M most Probable Configuration in Bayesian Networks. Statistics and Computing 2, 159–173 (1998)
Ochoa, A.: Linear Entropic Mutation (submitted for publication) (2009)
Ochoa, A.: The Random Class of B-functions (unpublished) (2009)
Ochoa, A., Soto, M.: Partial Evaluation in Genetic Algorithms. In: IEA/AIE 1997: Proceedings of the 10th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, pp. 217–222. Goose Pond Press, Atlanta (1997)
Ochoa, A., Soto, M.: Linking Entropy to Estimation of Distribution Algorithms. In: Lozano, J., Larrañaga, P., Inza, I., Bengoetxea, E. (eds.) Towards a new Evolutionary Computation. Advances on Estimation of Distribution Algorithms. Studies in Fuzziness and Soft Computing, pp. 1–38. Springer, Heidelberg (2006)
Ochoa, A., Soto, M.: On the Performance of the Bayesian Optimization Algorithm with B-functions. Tech. Rep. 383, ICIMAF (2006)
Ochoa, A., Höns, R., Soto, M., Müehlenbein, H.: A Maximum Entropy Approach to Sampling in EDA-the Single Connected case. In: Sanfeliu, A., Ruiz-Shulcloper, J. (eds.) CIARP 2003. LNCS, vol. 2905, pp. 683–690. Springer, Heidelberg (2003)
Ong, Y.S., Nair, P.B., Keane, A.J.: Evolutionary optimization of computationally expensive problems via surrogate modeling. American Institute of Aeronautics and Astronautics Journal 41(4), 687–696 (2003)
Pelikan, M., Goldberg, D.E., Cantú-Paz, E.: BOA: The Bayesian optimization algorithm. In: Banzhaf, W., Daida, J., Eiben, A.E., Garzon, M.H., var, V.H., Jakiela, M., Smith, R.E. (eds.) Proceedings of the Genetic and Evolutionary Computation Conference GECCO 1999, Orlando, FL, vol. 1, pp. 525–532. Morgan Kaufmann Publishers, San Francisco (1999)
Pelikan, M., Sastry, K., Cant-Paz, E. (eds.): Scalable Optimization via Probabilistic Modeling. Studies in Computational Intelligence, vol. 33. Springer, Heidelberg (2006)
Sastry, K., Pelikan, M., Goldberg, D.E.: Efficiency Enhancement of Estimation of Distribution Algorithms. In: Pelikan, M., Sastry, K., Cant-Paz, E. (eds.) Scalable Optimization via Probabilistic Modeling. Studies in Computational Intelligence, vol. 33, pp. 161–186. Springer, Heidelberg (2006)
Schäfer, J., Strimmer, K.: A Shrinkage Approach to Large-Scale Covariance Matrix Estimation and Implications for Functional Genomics. Statistical Applications in Genetics and Molecular Biology 4(1) (2005)
Shannon, C.E.: A Mathematical Theory of Communication. Bell System Technical Journal 27, 379–423 (1948)
Smith, R.E., Dike, B., Stegmann, S.: Fitness inheritance in genetic algorithms. In: Proceedings of the 1995 ACM Symposium on Applied Computing, SAC 1995, pp. 345–350 (1995)
Soto, M.: Un estudio sobre los algoritmos evolutivos basados en redes bayesianas simplemente conectadas y su costo de evaluación. PhD thesis, Instituto de Cibernética, Matemática y Física, La Habana (in Spanish). Adviser: A. Ochoa (2003)
Soto, M., Ochoa, A.: A Factorized Distribution Algorithm based on polytrees. In: Congress on Evolutionary Computation, CEC 2000, California, pp. 232–237 (2000)
Soto, M., Ochoa, A., Rodríguez-Ojea, L.: An Empirical Stuty on Oversampled Selection in Estimation of Distribution Algorithms. Tech. Rep. 494, ICIMAF (2008)
Yanover, C., Weiss, Y.: Finding the M most Probable Configurations in Arbitrary Graphical Models. In: Thrun, S., Saul, L., Schollkopf, B. (eds.) Advances in Neural Information Processing Systems, vol. 16. MIT Press, Cambridge (2004)
Yunpeng, C., Xiaomin, S., Peifa, J.: Probabilistic Modeling for Continuos EDA with Boltzmann Selection and Kullback-Liebler Divergence. In: Proceedings of GECCO Conference (2006)
Zhou, Z.Z., Ong, Y.S., Nair, P.B., Keane, A.J., Lum, K.Y.: Combining global and local surrogate models to accelerate evolutionary optimization. IEEE Transactions on Systems, Man and Cybernetics-Part C 37(1), 66–76 (2007)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Ochoa, A. (2010). Opportunities for Expensive Optimization with Estimation of Distribution Algorithms. In: Tenne, Y., Goh, CK. (eds) Computational Intelligence in Expensive Optimization Problems. Adaptation Learning and Optimization, vol 2. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-10701-6_8
Download citation
DOI: https://doi.org/10.1007/978-3-642-10701-6_8
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-10700-9
Online ISBN: 978-3-642-10701-6
eBook Packages: EngineeringEngineering (R0)