Skip to main content
SpringerLink
Log in

Evolutionary optimization using graphical models

  • Special Issue
  • Published:
New Generation Computing Aims and scope Submit manuscript

Abstract

We have previously shown that a genetic algorithm can be approximated by an evolutionary algorithm using the product of univariate marginal distributions of selected points as search distribution. This algorithm (UMDA) successfully optimizes difficult multi-modal optimization problems. For correlated fitness landscapes more complex factorizations of the search distribution have to be used. These factorizations are used by the Factorized Distribution Algorithm FDA. In this paper we extend FDA to an algorithm which computes a factorization from the data. The factorization can be represented by a Bayesian network. The Bayesian network is used to generate the search points.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

References

  1. Bouckaert, R.R., “Properties of Bayesian Network Learning Algorithms,” R. Lopez de Mantaras and D. Poole, (Eds.)Proc. Tenth Conference on Uncertainty in Artificial Intelligence, Morgan Kaufmann, San Francisco, pp. 102–0109, 1994.

  2. Baluja, S. and Davies, S., “Using Optimal Dependency-Trees for Combinatorial Optimization: Learning the Structure of the Search Space,”Technical report CMU-CS-97-107, Carnegie-Mellon University, Pittsburgh, 1997.

    Google Scholar 

  3. De Bonet, J.S., Isbell, Ch. L. and Viola, P., “MIMIC: Finding Optima by Estimating Probability Densities,” Mozer, M., Jordan, M. and Petsche, Th. (Eds)Advances in Neural Information Processing Systems 9, pp. 424–431, 1997.

  4. Goldberg, D. E.,Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley, Reading, 1989.

  5. Harik, G.,Linkage Learning via Probabilistic Modeling, the ECGA IlliGal Technical Report 99010, University of Illinois at Urbana-Champaign, 1999.

  6. Jordan, M. I. (ed.),Learning in Graphical Models, MIT Press, Cambridge, 1999.

    Google Scholar 

  7. Mühlenbein, H., “The Equation for Response to Selection and its Use for Prediction,”Evolutionary Computation, 5, pp. 303–346, 1997.

    Article  Google Scholar 

  8. Mühlenbein, H., Mahnig, Th. and Ochoa, R., “Schemata, Distributions and Graphical Models in Evolutionary Optimization,”Journal of Heuristics, 5, pp. 215–247, 1999.

    Article  MATH  Google Scholar 

  9. Mühlenbein, H. and Mahnig, Th., “Convergence Theory and Applications of the Factorized Distribution Algorithm,”Journal of Computing and Information Technology, 7, pp. 19–32, 1999.

    Google Scholar 

  10. Mühlenbein, H. and Mahnig, Th., “FDA — A Scalable Evolutionary Algorithm for the Optimization of Additively Decomposed Discrete Functions,”to appear in Evolutionary Computation 7.4, 1999.

  11. Pelikan, M., Goldberg, D. E. and Cantu-Paz, E.,BOA: The Bayesian Optimization Algorithm, IlliGAL Technical Report 99003, University of Illinois at Urbana-Champagin, 1999.

  12. Schwarz, G, “Estimating the Dimension of a Model,”Annals of Statistics, 7, pp. 461–464, 1978.

    Article  Google Scholar 

  13. Zhang, B.-T., Ohm, P. and Mühlenbein, H., “Evolutionary Induction of Sparse Neural Trees,”Evolutionary Computation, 5, pp. 213–236, 1997.

    Article  MATH  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Theoretical Foundation GMD Lab. Real World Computing Partnership, GMD FZ Informationstechnik, 53754, St. Augustin

    Heinz Mühlenbein & Thilo Mahnig

Authors
  1. Heinz Mühlenbein
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Thilo Mahnig
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Heinz Mühlenbein.

Additional information

Heinz Mühlenbein, Ph.D.: He is a research manager at GMD, the German national center for information technology. He obtained his diploma in mathematics from the University of Cologne in 1969, and his Ph.D from the University of Bonn in 1975. He entered GMD in 1969. He has worked on performance analysis of computer systems, computer networks, and massively parallel computers. Since 1988 his research interests are in Natural Computation. He was Visiting Professor at the Universities Paderborn, Bonn, Edinburgh and Carnegie-Mellon University. He has published over 60 research papers. He initiated the international conference series in natural computation PPSN in 1990. From 1993 to 1998 he was responsible European editor of Evolutionary Computation. He is presently on the Editorial Board of Evolutionary Computation, Scientific Computation and Journal of Heuristics.

Thilo Mahnig, Ph.D. student: He is working at GMD — German National Research Center for Information Technology in St. Augustin. He obtained his diploma in mathematics from the University of Bonn in differential geometry in 1996. His research interest lies in the theory of population based optimization algorithms. He has co-authored several papers with Heinz Mühlenbein.

About this article

Cite this article

Mühlenbein, H., Mahnig, T. Evolutionary optimization using graphical models. New Gener Comput 18, 157–166 (2000). https://doi.org/10.1007/BF03037594

Download citation

  • Received:

  • Revised:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF03037594

Keywords

Search

Navigation