Decreasing the number of evaluations in evolutionary algorithms by using a meta-model of the fitness function

Authors:
Jens Ziegler;Wolfgang Banzhaf
Affiliations:
University of Dortmund, Department of Computer Science, Dortmund, Germany;University of Dortmund, Department of Computer Science, Dortmund, Germany
Venue:
EuroGP'03 Proceedings of the 6th European conference on Genetic programming
Year:
2003

Citing 6
Cited 2

Genetic programming: an introduction: on the automatic evolution of computer programs and its applications

Genetic programming: an introduction: on the automatic evolution of computer programs and its applications
Noise and the Reality Gap: The Use of Simulation in Evolutionary Robotics

Proceedings of the Third European Conference on Advances in Artificial Life
Limiting the Number of Fitness Cases in Genetic Programming Using Statistics

PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Metamodel-Assisted Evolution Strategies

PPSN VII Proceedings of the 7th International Conference on Parallel Problem Solving from Nature
Fitness Approximation In Evolutionary Computation - a Survey

GECCO '02 Proceedings of the Genetic and Evolutionary Computation Conference
A framework for evolutionary optimization with approximate fitnessfunctions

IEEE Transactions on Evolutionary Computation

Implementation issues for an interactive evolutionary computation system

Proceedings of the 9th annual conference companion on Genetic and evolutionary computation
Population clustering in genetic programming

EuroGP'06 Proceedings of the 9th European conference on Genetic Programming

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper a method is presented that decreases the necessary number of evaluations in Evolutionary Algorithms. A classifier with confidence information is evolved to replace time consuming evaluations during tournament selection. Experimental analysis of a mathematical example and the application of the method to the problem of evolving walking patterns for quadruped robots show the potential of the presented approach.