An evolutionary function approximation approach to compute prediction in XCSF

Authors:
Ali Hamzeh;Adel Rahmani
Affiliations:
Department of Computer Engineering, Iran University of Science and Technology, Tehran, Iran;Department of Computer Engineering, Iran University of Science and Technology, Tehran, Iran
Venue:
ECML'05 Proceedings of the 16th European conference on Machine Learning
Year:
2005

Citing 6
Cited 6

Adaptive switching circuits

Neurocomputing: foundations of research
Adaptation in natural and artificial systems

Adaptation in natural and artificial systems
Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
Classifiers that approximate functions

Natural Computing: an international journal
Classifier fitness based on accuracy

Evolutionary Computation
Tournament selection: stable fitness pressure in XCS

GECCO'03 Proceedings of the 2003 international conference on Genetic and evolutionary computation: PartII

Learning classifier systems: a complete introduction, review, and roadmap

Journal of Artificial Evolution and Applications
Approximating Arbitrary Reinforcement Signal by Learning Classifier Systems using Micro Genetic Algorithm

Fundamenta Informaticae
Learning classifier systems: a complete introduction, review, and roadmap

Journal of Artificial Evolution and Applications
Three architectures for continuous action

IWLCS'03-05 Proceedings of the 2003-2005 international conference on Learning classifier systems
A pattern based evolutionary approach to prediction computation in XCSF

ICNC'06 Proceedings of the Second international conference on Advances in Natural Computation - Volume Part I
Approximating Arbitrary Reinforcement Signal by Learning Classifier Systems using Micro Genetic Algorithm

Fundamenta Informaticae

Quantified Score

Hi-index	0.00

Visualization

Abstract

XCSF is a new extension to XCS that is developed to extend XCS's reward calculation capability via computing. This new feature is called computable prediction. The first version of XCSF tries to find the most appropriate equation to compute each classifier's reward using a weight update mechanism. In this paper, we try to propose a new evolutionary mechanism to compute these equations using genetic algorithms.