Alexander E.I. Brownlee
Martin Pelikan
ABSTRACT
Learning a good model structure is important to the efficient solving of problems by estimation of distribution algorithms. In this paper we present the results of a series of experiments, applying a structure learning algorithm for undirected probabilistic graphical models based on statistical dependency tests to three fitness functions with different selection operators, proportions and pressures. The number of spurious interactions found by the algorithm are measured and reported. Truncation selection, and its complement (selecting only low fitness solutions) prove quite robust, resulting in a similar number of spurious dependencies regardless of selection pressure. In contrast, tournament and fitness proportionate selection are strongly affected by the selection proportion and pressure.
- S. Baluja and S. Davies. Using optimal dependency-trees for combinational optimization. In ICML '97: Proceedings of the Fourteenth International Conference on Machine Learning, pages 30--38. Morgan Kaufmann Publishers Inc, 1997. Google Scholar
Digital Library
- S. Boslaugh and P. A. Watters. Statistics: A Desktop Quick Reference. O'Reilly, 2008.Google Scholar
- A. E. I. Brownlee. Multivariate Markov Networks for Fitness Modelling in an Estimation of Distribution Algorithm. PhD thesis, Robert Gordon University, Aberdeen, May 2009.Google Scholar
- A. E. I. Brownlee, J. A. W. McCall, and M. Pelikan. Influence of selection on structure learning in Markov network EDAs: An empirical study. Technical Report MEDAL Report No. 2012006, Missouri Estimation of Distribution Algorithms Laboratory, Univ. of Missouri in St. Louis, 2012. Online: http://medal-lab.org/files/2012006.pdf.Google ScholarDigital Library
- A. E. I. Brownlee, J. A. W. McCall, S. K. Shakya, and Q. Zhang. Structure Learning and Optimisation in a Markov-network based Estimation of Distribution Algorithm. In Proceedings of the IEEE Congress on Evolutionary Computation (CEC 2009), pages 447--454, Trondheim, Norway, 2009. IEEE Press. Google ScholarDigital Library
- A. E. I. Brownlee, J. A. W. McCall, Q. Zhang, and D. Brown. Approaches to Selection and their effect on Fitness Modeling in an Estimation of Distribution Algorithm. In Proc. of the IEEE World Congress on Computational Intelligence (CEC 2008), pages 2621--2628, Hong Kong, China, 2008. IEEE Press.Google ScholarCross Ref
- J. S. de Bonet, C. L. Isbell Jr., and P. Viola. MIMIC: Finding optima by estimating probability densities. In M. C. Mozer, M. I. Jordan, and T. Petsche, editors, Advances in Neural Information Processing Systems, 1997.Google Scholar
- C. Echegoyen, J. Lozano, R. Santana, and P. Larranaga. Exact Bayesian network learning in estimation of distribution algorithms. In Proceedings of the IEEE Congress on Evolutionary Computation (CEC 2007), pages 1051--1058, sept. 2007.Google ScholarCross Ref
- M. Hauschild and M. Pelikan. An introduction and survey of estimation of distribution algorithms. Swarm and Evolutionary Computation, 1(3):111 -- 128, 2011.Google ScholarCross Ref
- M. Hauschild, M. Pelikan, K. Sastry, and C. Lima. Analyzing probabilistic models in hierarchical BOA. IEEE Transactions on Evolutionary Computation, 13:1199--1217, December 2009. Google ScholarDigital Library
- Y. Hong, G. Zhu, S. Kwong, and Q. Ren. Estimation of distribution algorithms making use of both high quality and low quality individuals. In IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2009), pages 1806--1813, August 2009.Google ScholarCross Ref
- L. Kallel, B. Naudts, and R. Reeves. Properties of fitness functions and search landscapes. In L. Kallel, B. Naudts, and A. Rogers, editors, Theoretical Aspects of Evolutionary Computing, pages 177--208. Springer Verlag, 2000. Google ScholarDigital Library
- P. Larranaga and J. A. Lozano. Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation. Kluwer Academic Publishers, Boston, 2002. Google ScholarDigital Library
- X. Li, S. Mabu, and K. Hirasawa. Use of infeasible individuals in probabilistic model building genetic network programming. In Proc. of the 13th Genetic and Evolutionary Computation COnference (GECCO 2011), pages 601--608, New York, NY, USA, 2011. ACM. Google ScholarDigital Library
- C. F. Lima, F. G. Lobo, M. Pelikan, and D. E. Goldberg. Model accuracy in the Bayesian optimization algorithm. Soft Computing, 15(7):1351--1371, 2010. Google ScholarDigital Library
- J. A. Lozano, P. Larranaga, I. Inza, and E. Bengoetxea. Towards a New Evolutionary Computation: Advances on Estimation of Distribution Algorithms (Studies in Fuzziness and Soft Computing). Springer-Verlag, 2006. Google ScholarDigital Library
- L. A. Marascuilo and M. McSweeney. Nonparametric and Distribution-Free Methods for Social Sciences. Brooks / Cole Publishing Company, California, 1977.Google Scholar
- T. Miquélez, E. Bengoetxea, and P. Larranaga. Evolutionary computation based on Bayesian classifiers. Int'l Jnl. of Applied Mathematics and Computer Science, 14(3):101--115, 2004.Google Scholar
- H. Mühlenbein and T. Mahnig. Evolutionary optimization using graphical models. New Gen. Comput., 18(2):157--166, 2000. Google ScholarDigital Library
- H. Mühlenbein and D. Schlierkamp-Voosen. Predictive models for the breeder genetic algorithm I. continuous parameter optimization. Evolutionary Computuation, 1(1):25--49, 1993. Google ScholarDigital Library
- M. Munetomo, N. Murao, and K. Akama. Introducing assignment functions to Bayesian optimization algorithms. Info. Sciences, 178(1):152 -- 163, 2008. Google ScholarDigital Library
- M. Pelikan. Bayesian optimization algorithm: from single level to hierarchy. PhD thesis, University of Illinois at Urbana-Champaign, Urbana, IL, 2002. Google ScholarDigital Library
- M. Pelikan and H. Mühlenbein. The bivariate marginal distribution algorithm. In R. Roy, T. Furuhashi, and P. K. Chawdhry, editors, Advances in Soft Computing - Engineering Design and Manufacturing, pages 521--535, 1999.Google Scholar
- M. Pelikan, K. Sastry, and D. E. Goldberg. Scalability of the Bayesian optimization algorithm. Int'l. Jnl. of Approximate Reasoning, 31(3):221 -- 258, 2002.Google ScholarCross Ref
- E. Radetic and M. Pelikan. Spurious dependencies and EDA scalability. In Genetic and Evolutionary Computation Conference, pages 303--310, 2010. Google ScholarDigital Library
- R. Santana. A Markov network based factorized distribution algorithm for optimization. In Proceedings of the 14th European Conference on Machine Learning; Lecture Notes in Artificial Intelligence, volume 2837, pages 337--348, Berlin, 2003. Springer-Verlag.Google ScholarDigital Library
- R. Santana. Estimation of distribution algorithms with Kikuchi approximations. Evolutionary Computation, 13(1):67--97, 2005. Google ScholarDigital Library
- R. Santana, C. Bielza, J. A. Lozano, and P. Larranaga. Mining probabilistic models learned by EDAs in the optimization of multi-objective problems. In Proceedings of the 11th Annual conference on Genetic and evolutionary computation (GECCO 2009), pages 445--452, New York, NY, USA, 2009. ACM. Google ScholarDigital Library
- R. Santana, P. Larranaga, and J. A. Lozano. Interactions and dependencies in estimation of distribution algorithms. In Proc. of the IEEE Congress on Evolutionary Computation (CEC 2005), volume 1, pages 1418--1425. IEEE Press, 2--4 Sept. 2005.Google ScholarCross Ref
- R. Santana, P. Larranaga, and J. A. Lozano. Challenges and open problems in discrete EDAs. Technical Report EHU-KZAA-IK-1/07, Department of Computer Science and Artificial Intelligence, University of the Basque Country, October 2007.Google Scholar
- R. Santana, P. Larranaga, and J. A. Lozano. Research topics on discrete estimation of distribution algorithms. Memetic Computing, 1(1):35--54, 2009.Google ScholarCross Ref
- S. Shakya and R. Santana. An EDA based on local Markov property and Gibbs sampling. In Proceedings of the Genetic and Evolutionary Computation COnference (GECCO 2008), pages 475--476. ACM Press, 2008. Google ScholarDigital Library
- S. Shakya and R. Santana, editors. Markov Networks in Evolutionary Computation. Springer, 2012.Google ScholarDigital Library
- S. K. Shakya, A. E. I. Brownlee, J. A. W. McCall, F. Fournier, and G. Owusu. A fully multivariate DEUM algorithm. In Proceedings of the IEEE Congress on Evolutionary Computation (CEC 2009), pages 479--486. IEEE Press, 2009. Google ScholarDigital Library
- S. K. Shakya and J. A. W. McCall. Optimization by estimation of distribution with DEUM framework based on Markov random fields. International Journal of Automation and Computing, 4(3):262--272, 2007.Google ScholarCross Ref
- D. Wallin and C. Ryan. Using over-sampling in a Bayesian classifier EDA to solve deceptive and hierarchical problems. In Proc. of the IEEE Congress on Evolutionary Computation, pages 1660--1667, 2009. Google ScholarDigital Library
- T.-L. Yu, K. Sastry, and D. E. Goldberg. Linkage learning, overlapping building blocks, and systematic strategy for scalable recombination. In Proceedings of the 2005 conference on Genetic and evolutionary computation, GECCO '05, pages 1217--1224, New York, NY, USA, 2005. ACM. Google ScholarDigital Library
- T.-L. Yu, K. Sastry, D. E. Goldberg, and M. Pelikan. Population sizing for entropy-based model building in discrete estimation of distribution algorithms. In Proceedings of the 9th annual conference on Genetic and evolutionary computation, GECCO '07, pages 601--608, New York, NY, USA, 2007. ACM. Google ScholarDigital Library
Index Terms
- Influence of selection on structure learning in markov network EDAs: an empirical study
Recommendations
The IBMAP approach for Markov network structure learning
In this work we consider the problem of learning the structure of Markov networks from data. We present an approach for tackling this problem called IBMAP, together with an efficient instantiation of the approach: the IBMAP-HC algorithm, designed for ...
A niching scheme for EDAs to reduce spurious dependencies
GECCO '13: Proceedings of the 15th annual conference on Genetic and evolutionary computationThis paper proposes a niching scheme, the dependency structure matrix restricted tournament replacement (DSMRTR). The restricted tournament replacement (RTR) is a well-known niching scheme in the field of estimation of distribution algorithms (EDAs). ...
An extended mutation concept for the local selection based differential evolution algorithm
GECCO '07: Proceedings of the 9th annual conference on Genetic and evolutionary computationA new mutation concept is proposed to generalize local selection based Differential Evolution algorithm to work in general multi-modal problems. Three variations of the proposed method are compared with classic Differential Evolution algorithm using a ...
Comments