Horizontal gene transfer for recombining graphs

• Published in: Genetic programming and evolvable machines Vol. 21; no. 3; pp. 321 - 347
• Main Authors: Atkinson, Timothy, Plump, Detlef, Stepney, Susan
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.09.2020; Springer Nature B.V
• Subjects: Artificial Intelligence; Biomedical Engineering and Bioengineering; Cartesian coordinates; Children; Compilers; Computer Science; Electrical Engineering; Evolutionary algorithms; Genetic algorithms; Graphs; Interpreters; Parents; Performance enhancement; Programming; Programming Languages; Programming Techniques; Software Engineering/Programming and Operating Systems; Special Issue: Highlights of Genetic Programming 2019 Events; Neuroevolution; Graph-based genetic programming; Horizontal gene transfer
• ISSN: 1389-2576; 1573-7632
• DOI: 10.1007/s10710-020-09378-1

We introduce a form of neutral horizontal gene transfer (HGT) to evolving graphs by graph programming (EGGP). We introduce the μ × λ evolutionary algorithm (EA), where μ parents each produce λ children who compete only with their parents. HGT events then copy the entire active component of one surviving parent into the inactive component of another parent, exchanging genetic information without reproduction. Experimental results from symbolic regression problems show that the introduction of the μ × λ EA and HGT events improve the performance of EGGP. Comparisons with genetic programming and Cartesian genetic programming strongly favour our proposed approach. We also investigate the effect of using HGT events in neuroevolution tasks. We again find that the introduction of HGT improves the performance of EGGP, demonstrating that HGT is an effective cross-domain mechanism for recombining graphs.