Journal of the Italian Statistical Society

, Volume 7, Issue 2, pp 197–208 | Cite as

A genetic algorithm for graphical model selection

  • Irene Poli
  • Alberto Roverato


Graphical log-linear model search is usually performed by using stepwise procedures in which edges are sequentially added or eliminated from the independence graph. In this paper we implement the search procedure as a genetic algorithm and propose a crossover operator which operates on subgraphs. In a simulation study the proposed procedure is shown to perform better than an automatic backward elimination procedure at the cost of a small increase of computational time.


AIC Genetic Algorithm Graphical model Log-linear model Model selection Undirected graph 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Akaike, H. (1973), Information theory and an extension of the maximum likelihood principle. In2nd International Symposium on Information Theory, (eds. B. N. Petrov and F. Csaki), 267–281. Budapest: Akademia Kiado.Google Scholar
  2. Badsberg, J. H. (1991), A guide to CoCo.Technical Report 91-43, Institute for Electronic Systems, University of Aalborg.Google Scholar
  3. Baker, J. (1987), Reducing bias and inefficiency in the selection algorithm. InGenetic Algorithms and Their Applications: Proceedings of the Second International Conference on Genetic Algorithms (ed. J. J. Grefenstette), Erlbaum.Google Scholar
  4. Bishop, Y. M., Fienberg, S. andHolland, P. (1975),Discrete Multivariate Analysis, Cambridge, Mass.: M.I.T. Press.MATHGoogle Scholar
  5. Box, G. E. P. (1957), Evolutionary operation: A method for increasing industrial productivity.J. Roy. Statist. Soc., C 6, 2, 81–101.Google Scholar
  6. Cowell, R. G., Dawid, A. P., Lauritzen, S. L. andSpiegelhalter, D. J. (1999),Probabilistic Networks and Expert Systems. New York: Springer-Verlag.MATHGoogle Scholar
  7. Darroch, J. N., Lauritzen, S. L. andSpeed, T. P. (1980), Markov fields and log-linear interaction models for contingency tables.Ann. Statist., 11, 3, 724–738.CrossRefMathSciNetGoogle Scholar
  8. Dawid, A. P., Lauritzen, S. L. (1993), Hyper Markov laws in the statistical analysis of decomposable graphical models.Ann. Statist., 21, 1272–1317.MATHCrossRefMathSciNetGoogle Scholar
  9. Edwards, D. E. (1995),Introduction to Graphical Modelling. New York: Springer-Verlag.MATHGoogle Scholar
  10. Edwards, D. E. andHavránek, T. (1987), A Fast Model Selection Procedure for Large Families of Models.J. Amer. Statist. Assoc., 82, 205–213.MATHCrossRefMathSciNetGoogle Scholar
  11. Goldberg, D. E. (1989),Genetic Algorithms in Search, Optimisation, and Machine Learning. Addison-Wesley.Google Scholar
  12. Holland, J. H. (1975),Adaptation in Natural and Artificial Systems, Ann Arbor: The University of Michigan Press.Google Scholar
  13. Holland, J. H. (1995),Hidden Order, how adaptation builds complexity. Reading: Addison-Wesley Publishing Corporation, Inc.Google Scholar
  14. Larrañaga, P., Murga, R. H., Poza, M. andKuijpers, C. M. H. (1996), Structure learning of Bayesian networks by hybrid genetic algorithms. InLearning from Data: Artificial Intelligence and Statistics V, (eds. D. Fisher and H. Lenz), 165–174. New York: Springer-Verlag.Google Scholar
  15. Lauritzen, S. L. (1996),Graphical Models. Oxford: Oxford University Press.Google Scholar
  16. Malvestuto, F. M. (1996), An axiomatisation of log-linear models with an application to the model-search problem. InLearning from Data: Artificial Intelligence and Statistics V, (eds. D. Fisher and H. Lenz), 175–184. New York: Springer-Verlag.Google Scholar
  17. Mitchell, M. (1996),An Introduction to Genetic Algorithms. M.I.T. Press.Google Scholar
  18. Whittaker, J. (1990),Graphical Models in Applied Multivariate Statistics. Chichester: Whiley.MATHGoogle Scholar

Copyright information

© Società Italiana di Statistica 1998

Authors and Affiliations

  1. 1.Università «Ca’ Foscari» di VeneziaItaly
  2. 2.Dipartimento di Economia PoliticaUniversità di Modena e Reggio EmiliaModenaItaly

Personalised recommendations