The detrimentality of crossover

Authors:
Andrew Czarn;Cara MacNish;Kaipillil Vijayan;Berwin Turlach
Affiliations:
School of Computer Science and Software Engineering, The University of Western Australia, Crawley, WA;School of Computer Science and Software Engineering, The University of Western Australia, Crawley, WA;School of Mathematics and Statistics, The University of Western Australia, Crawley, WA;School of Mathematics and Statistics, The University of Western Australia, Crawley, WA
Venue:
AI'07 Proceedings of the 20th Australian joint conference on Advances in artificial intelligence
Year:
2007

Citing 6
Cited 1

On the analysis of the (1+ 1) evolutionary algorithm

Theoretical Computer Science
A Study of Control Parameters Affecting Online Performance of Genetic Algorithms for Function Optimization

Proceedings of the 3rd International Conference on Genetic Algorithms
Crossover's Niche

Proceedings of the 5th International Conference on Genetic Algorithms
An analysis of the behavior of a class of genetic adaptive systems.

An analysis of the behavior of a class of genetic adaptive systems.
Benchmarking evolutionary and hybrid algorithms using randomized self-similar landscapes

SEAL'06 Proceedings of the 6th international conference on Simulated Evolution And Learning
Statistical exploratory analysis of genetic algorithms

IEEE Transactions on Evolutionary Computation

A survey of techniques for characterising fitness landscapes and some possible ways forward

Information Sciences: an International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

The traditional concept of a genetic algorithm (GA) is that of selection, crossover and mutation. However, data from the literature has suggested that the niche for the beneficial effect of crossover upon GA performance may be smaller than has been traditionally held. We explored the class of problems for which crossover is detrimental by performing a statistical analysis of two test problem suites, one comprising linear-separable non-rotated functions and the other comprising the same functions rotated by 45 degrees rendering them not-linear-separable We find that for the difficult rotated functions the crossover operator is detrimental to the performance of the GA. We conjecture that what makes a problem difficult for the GA is complex and involves factors such as the degree of optimization at local minima due to crossover, the bias associated with the mutation operator and the Hamming Distances present in the individual problems due to the encoding. Finally, we test our GA on a practical landscape minimization problem to see if the results obtained match those from the difficult rotated functions. We find that they match and that the features which make certain of the test functions difficult are also present in the real world problem.