A Comparative Study on Crossover in Cartesian Genetic Programming

Husa, Jakub; Kalkreuth, Roman

doi:10.1007/978-3-319-77553-1_13

Cited by 25 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use the 1 + EA with = 4 . We do not use any of the published CGP crossover operators, as their usefulness, particularly on symbolic regression problems, remains disputed [15], and [24,37] recommend the 1 +…”

Section: Experimental Settingsmentioning

confidence: 99%

“…A number of crossover operators have been used in CGP, including uniform crossover [23], arithmetic crossover on a vector representation [5], and subgraph crossover [17]. Empirical comparison [15] shows that these crossover operators do not always aid performance, and that CGP with mutation only can sometimes be the best performing approach. Current advice [24] is that the 'standard' CGP approach remains to use mutation only.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Horizontal gene transfer for recombining graphs

Atkinson

Plump

Stepney

2020

Genet Program Evolvable Mach

View full text Add to dashboard Cite

We introduce a form of neutral horizontal gene transfer (HGT) to evolving graphs by graph programming (EGGP). We introduce the × evolutionary algorithm (EA), where parents each produce 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 × 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.

show abstract

Section: Experimental Settingsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Horizontal gene transfer for recombining graphs

Atkinson

Plump

Stepney

2020

Genet Program Evolvable Mach

View full text Add to dashboard Cite

show abstract

“…Kalkreuth and Husa have recently proposed block crossover [32]. This is defined using the one-dimensional representation of CGP.…”

Section: Crossovermentioning

confidence: 99%

Cartesian genetic programming: its status and future

Miller

2019

Genet Program Evolvable Mach

View full text Add to dashboard Cite

Cartesian genetic programming, a well-established method of genetic programming, is approximately 20 years old. It represents solutions to computational problems as graphs. Its genetic encoding includes explicitly redundant genes which are wellknown to assist in effective evolutionary search. In this article, we review and compare many of the important aspects of the method and findings discussed since its inception. In the process, we make many suggestions for further work which could improve the efficiency of the CGP for solving computational problems.

show abstract

“…We use the 1 + Ú EA with Ú = 4. We do not use any of the published CGP crossover operators, as their usefulness, particularly on symbolic regression problems, remains disputed [9], and [15,23] recommend the 1 + Ú approach. We also use no form of depth control with CGP, as the approach is known to have inherent anti-bloat biases [21].…”

Section: Experimental Settingsmentioning

confidence: 99%

“…A number of crossover operators have been used in CGP, including uniform crossover [14], arithmetic crossover on a vector representation [4], and subgraph crossover [10]. Empirical comparison [9] shows that these crossover operators do not always aid performance, and that CGP with mutation only can sometimes be the best performing approach. Current advice [15,23] is that the 'standard' CGP approach remains to use mutation only.…”

Section: Introductionmentioning

confidence: 99%

Evolving graphs with horizontal gene transfer

Atkinson

Plump

Stepney

2019

Proceedings of the Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

We introduce a form of neutral Horizontal Gene Transfer (HGT) to Evolving Graphs by Graph Programming (EGGP). We introduce the Û × Ú evolutionary algorithm, where Û parents each produce Ú children who compete with only 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 14 symbolic regression benchmark problems show that the introduction of the Û×Ú EA and HGT events improve the performance of EGGP. Comparisons with Genetic Programming and Cartesian Genetic Programming strongly favour our proposed approach. CCS CONCEPTS• Computing methodologies → Genetic programming;

show abstract

A Comparative Study on Crossover in Cartesian Genetic Programming

Cited by 25 publications

References 15 publications

Horizontal gene transfer for recombining graphs

Horizontal gene transfer for recombining graphs

Cartesian genetic programming: its status and future

Evolving graphs with horizontal gene transfer

Contact Info

Product

Resources

About