Skip to main content

Genetic Programming Crossover: Does It Cross Over?

Johnson, Colin G. (2009) Genetic Programming Crossover: Does It Cross Over? In: EuroGP 2009: Genetic Programming. Lecture Notes in Computer Science . pp. 97-108. Springer, Heidelberg ISBN 978-3-642-01180-1. E-ISBN 978-3-642-01181-8. (doi:10.1007/978-3-642-01181-8_9)

PDF - Author's Accepted Manuscript
Download (282kB) Preview
[img]
Preview
Official URL
https://doi.org/10.1007/978-3-642-01181-8_9

Abstract

One justification for the use of crossover operators in Genetic Programming is that the crossover of program syntax gives rise to the crossover of information at the semantic level. In particular, a fitness-increasing crossover is presumed to act by combining fitness-contributing components of both parents. In this paper we investigate a particular interpretation of this hypothesis via an experimental study of 70 GP runs, in which we categorise each crossover event by its fitness properties and the information that contributes most strongly to those fitness properties. Some tentative evidence in support of the above hypothesis is extracted from this categorisation.

Item Type: Conference or workshop item (Proceeding)
DOI/Identification number: 10.1007/978-3-642-01181-8_9
Subjects: Q Science > QA Mathematics (inc Computing science) > QA 76 Software, computer programming,
Divisions: Faculties > Sciences > School of Computing
Depositing User: Colin Johnson
Date Deposited: 14 Dec 2018 10:23 UTC
Last Modified: 30 May 2019 08:34 UTC
Resource URI: https://kar.kent.ac.uk/id/eprint/71019 (The current URI for this page, for reference purposes)
Johnson, Colin G.: https://orcid.org/0000-0002-9236-6581
  • Depositors only (login required):

Downloads

Downloads per month over past year