reCORE --- a coevolutionary algorithm for rule extraction

Authors:
Bogdan Trawiński;Grzegorz Matoga
Affiliations:
Instititute of Informatics, Wroclaw University of Technology, Wrocław, Poland;Instititute of Informatics, Wroclaw University of Technology, Wrocław, Poland
Venue:
SIDE'12 Proceedings of the 2012 international conference on Swarm and Evolutionary Computation
Year:
2012

Citing 11
Cited 0

A Knowledge-Intensive Genetic Algorithm for Supervised Learning

Machine Learning - Special issue on genetic algorithms
Competition-Based Induction of Decision Models from Examples

Machine Learning - Special issue on genetic algorithms
Outline for a Logical Theory of Adaptive Systems

Journal of the ACM (JACM)
Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control and Artificial Intelligence

Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control and Artificial Intelligence
Book Review: Data Mining Using Grammar-Based Genetic Programming and Applications

Genetic Programming and Evolvable Machines
SIA: A Supervised Inductive Algorithm with Genetic Search for Learning Attributes based Concepts

ECML '93 Proceedings of the European Conference on Machine Learning
Classifier fitness based on accuracy

Evolutionary Computation
A systematic analysis of performance measures for classification tasks

Information Processing and Management: an International Journal
Flexible learning of problem solving heuristics through adaptive search

IJCAI'83 Proceedings of the Eighth international joint conference on Artificial intelligence - Volume 1
The WEKA data mining software: an update

ACM SIGKDD Explorations Newsletter
Rule induction based-on coevolutionary algorithms for image annotation

ACIIDS'11 Proceedings of the Third international conference on Intelligent information and database systems - Volume Part II

Quantified Score

Hi-index	0.00

Visualization

Abstract

A coevolutionary algorithm called reCORE for rule extraction from databases was proposed in the paper. Knowledge is extracted in the form of simple implication rules IF-THEN-ELSE. There are two populations involved, one for rules and second for rule sets. Each population has a distinct evolution scheme. One individual from rule set population and a number of individuals from rule population contribute to final result. Populations are synchronized to keep the cross-references up to date. The concept of the algorithm and the results of comparative research are presented. A cross-validation is used for examination, and the final conclusions are drawn based upon statistically analyzed results. They state that, the effectiveness of the reCORE algorithm is comparable to other rule classifiers, such as Ridor or JRip.