Redundancy and computational efficiency in Cartesian genetic programming

J F Miller; S L Smith

doi:10.1109/TEVC.2006.871253

Redundancy and computational efficiency in Cartesian genetic programming

J F Miller, S L Smith

Electronic Engineering

Research output: Contribution to journal › Article › peer-review

Abstract

The graph-based Cartesian genetic programming system has an unusual genotype representation with a number of advantageous properties. It has a form of redundancy whose role has received little attention in the published literature. The representation has genes that can be activated or deactivated by mutation operators during evolution. It has been demonstrated that this "junk" has a useful role and is very beneficial in evolutionary search. The results presented demonstrate the role of mutation and genotype length in the evolvability of the representation. It is found that the most evolvable representations occur when the genotype is extremely large and in which over 95% of the genes are inactive.

Original language	English
Pages (from-to)	167-174
Number of pages	8
Journal	IEEE Transactions on Evolutionary Computation
Volume	10
Issue number	2
DOIs	https://doi.org/10.1109/TEVC.2006.871253
Publication status	Published - Apr 2006

Keywords

Cartesian genetic programming (CGP)
code bloat
genetic programming
graph-based representations
introns
NEUTRALITY
ADAPTATION
LANDSCAPE
EVOLUTION

Access to Document

10.1109/TEVC.2006.871253

http://dx.doi.org/10.1109/TEVC.2006.871253

Cite this

@article{689e4f85fec84c91a990544a0b0dfd61,

title = "Redundancy and computational efficiency in Cartesian genetic programming",

abstract = "The graph-based Cartesian genetic programming system has an unusual genotype representation with a number of advantageous properties. It has a form of redundancy whose role has received little attention in the published literature. The representation has genes that can be activated or deactivated by mutation operators during evolution. It has been demonstrated that this {"}junk{"} has a useful role and is very beneficial in evolutionary search. The results presented demonstrate the role of mutation and genotype length in the evolvability of the representation. It is found that the most evolvable representations occur when the genotype is extremely large and in which over 95% of the genes are inactive.",

keywords = "Cartesian genetic programming (CGP), code bloat, genetic programming, graph-based representations, introns, NEUTRALITY, ADAPTATION, LANDSCAPE, EVOLUTION",

author = "Miller, {J F} and Smith, {S L}",

year = "2006",

month = apr,

doi = "10.1109/TEVC.2006.871253",

language = "English",

volume = "10",

pages = "167--174",

journal = "IEEE Transactions on Evolutionary Computation",

issn = "1089-778X",

publisher = "IEEE",

number = "2",

}

TY - JOUR

T1 - Redundancy and computational efficiency in Cartesian genetic programming

AU - Miller, J F

AU - Smith, S L

PY - 2006/4

Y1 - 2006/4

N2 - The graph-based Cartesian genetic programming system has an unusual genotype representation with a number of advantageous properties. It has a form of redundancy whose role has received little attention in the published literature. The representation has genes that can be activated or deactivated by mutation operators during evolution. It has been demonstrated that this "junk" has a useful role and is very beneficial in evolutionary search. The results presented demonstrate the role of mutation and genotype length in the evolvability of the representation. It is found that the most evolvable representations occur when the genotype is extremely large and in which over 95% of the genes are inactive.

AB - The graph-based Cartesian genetic programming system has an unusual genotype representation with a number of advantageous properties. It has a form of redundancy whose role has received little attention in the published literature. The representation has genes that can be activated or deactivated by mutation operators during evolution. It has been demonstrated that this "junk" has a useful role and is very beneficial in evolutionary search. The results presented demonstrate the role of mutation and genotype length in the evolvability of the representation. It is found that the most evolvable representations occur when the genotype is extremely large and in which over 95% of the genes are inactive.

KW - Cartesian genetic programming (CGP)

KW - code bloat

KW - genetic programming

KW - graph-based representations

KW - introns

KW - NEUTRALITY

KW - ADAPTATION

KW - LANDSCAPE

KW - EVOLUTION

U2 - 10.1109/TEVC.2006.871253

DO - 10.1109/TEVC.2006.871253

M3 - Article

SN - 1089-778X

VL - 10

SP - 167

EP - 174

JO - IEEE Transactions on Evolutionary Computation

JF - IEEE Transactions on Evolutionary Computation

IS - 2

ER -