Abstract
The paper describes a learning environment named DEL-VAUX for classification tasks that learns Bayesian rule-sets from sets of examples. A genetic algorithm approach is used for this purpose, in which a population consists of sets of rule-sets that generate offspring through the exchange of rules, permitting fitter rule-sets to produce offspring with a higher probability. A bucket brigade algorithm for Bayesian rule-sets called reward punishment mechanism is introduced, which evaluates the performance of a Bayesian rule within a rule-set. It employs fuzzy techniques to measure the ”goodness” of a rule within a rule-set.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
De Jong, K.: ”Genetic Algorithm-based Learning” in Kodratoff (Eds.) et al.: ”Machine Learning”, Volume 3, Chapter 21, Morgan Kaufmann, San Mateo, 1990.
Eick, C. and Mehta, N.: ” Decision Making Involving Imperfect Knowledge”, IEEE Transactions on Systems, Man, and Cybernetics, pp. 840–850, June 1993.
Eick, C. and Jong, D.: ”Learning Bayesian Classification Rules Through Genetic Algorithms”, in Proc. Int. Conf. on Information and Knowledge Management, pp. 305–313, Washington, Nov. 1993.
Holland, J.: ”Escaping brittleness: the possibilities of general purpose learning algorithms applied to parallel rule-based systems”, in Michalski,R. et al.(Eds.): Machine Learning, an Arficial Intelligence Approach (Vol.2), Morgan Kaufmann, 1986.
Janikow, Cezary: ”A Knowledge-Intensive Genetic Algorithm for Supervised Learning”, Machine Learning, vol. 13, pp. 189–228, 1993.
Oliver, Jim: ”Discovering Individual Decision Rules: an Application of Genetic Algorithms”, in Proc. Fifth Int. Conference on Genetic Algorithms, pp. 216–222, Chicago, October 1993.
Toto, Erna: ”Combining Rule Debugging Techniques with Genetic Algorithms — a Hybrid Approach to Learn Rules for Classification Tasks”, Master's Thesis, Department of Computer Science, University of Houston, December 1993.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1994 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Eick, C.F., Toto, E. (1994). Evaluation and enhancement of Bayesian rule-sets in a genetic algorithm learning environment for classification tasks. In: Raś, Z.W., Zemankova, M. (eds) Methodologies for Intelligent Systems. ISMIS 1994. Lecture Notes in Computer Science, vol 869. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-58495-1_37
Download citation
DOI: https://doi.org/10.1007/3-540-58495-1_37
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-58495-7
Online ISBN: 978-3-540-49010-4
eBook Packages: Springer Book Archive