M. Heywood, and A. Zincir-Heywood. Systems, Man, and Cybernetics, 2000 IEEE International
Conference, 5, page 3823--3828. IEEE Press, (8-11 October 2000)
Abstract
Genetic programming arguably represents the most
general form of evolutionary computation. However, such
generality is not without significant computational
overheads. Particularly, the cost of evaluating the
fitness of individuals in any form of evolutionary
computation represents the single most significant
computational bottleneck. A less widely acknowledged
computational overhead in GP involves the
implementation of the crossover operator. To this end a
page-based definition of individuals is used to
restrict crossover to equal length code fragments.
Moreover, by using a register-machine context, the
significance of a priori internal register external
output definitions is emphasized.
%0 Conference Paper
%1 Heywood:2000:PBGP
%A Heywood, M. I.
%A Zincir-Heywood, A. N.
%B Systems, Man, and Cybernetics, 2000 IEEE International
Conference
%C IEEE Press
%D 2000
%K a algorithms, code computation, computational crossover definitions equal evolutionary external fitness fragments, genetic individuals, internal length linear of operator, output overheads, page-based priori programming, register register-machine,
%P 3823--3828
%T Page-based linear genetic programming
%U http://ieeexplore.ieee.org/iel5/7099/19140/00886606.pdf?isNumber=19140
%V 5
%X Genetic programming arguably represents the most
general form of evolutionary computation. However, such
generality is not without significant computational
overheads. Particularly, the cost of evaluating the
fitness of individuals in any form of evolutionary
computation represents the single most significant
computational bottleneck. A less widely acknowledged
computational overhead in GP involves the
implementation of the crossover operator. To this end a
page-based definition of individuals is used to
restrict crossover to equal length code fragments.
Moreover, by using a register-machine context, the
significance of a priori internal register external
output definitions is emphasized.
%@ 0-7803-6583-6
@inproceedings{Heywood:2000:PBGP,
abstract = {Genetic programming arguably represents the most
general form of evolutionary computation. However, such
generality is not without significant computational
overheads. Particularly, the cost of evaluating the
fitness of individuals in any form of evolutionary
computation represents the single most significant
computational bottleneck. A less widely acknowledged
computational overhead in GP involves the
implementation of the crossover operator. To this end a
page-based definition of individuals is used to
restrict crossover to equal length code fragments.
Moreover, by using a register-machine context, the
significance of a priori internal register external
output definitions is emphasized.},
added-at = {2008-06-19T17:35:00.000+0200},
address = {IEEE Press},
author = {Heywood, M. I. and Zincir-Heywood, A. N.},
biburl = {https://www.bibsonomy.org/bibtex/24c713e974de2ff377061d7071b937108/brazovayeye},
booktitle = {Systems, Man, and Cybernetics, 2000 IEEE International
Conference},
interhash = {3683a1563b5747e6133931439f49e1ba},
intrahash = {4c713e974de2ff377061d7071b937108},
isbn = {0-7803-6583-6},
keywords = {a algorithms, code computation, computational crossover definitions equal evolutionary external fitness fragments, genetic individuals, internal length linear of operator, output overheads, page-based priori programming, register register-machine,},
month = {8-11 October},
pages = {3823--3828},
size = {6 pages},
timestamp = {2008-06-19T17:41:24.000+0200},
title = {Page-based linear genetic programming},
url = {http://ieeexplore.ieee.org/iel5/7099/19140/00886606.pdf?isNumber=19140},
volume = 5,
year = 2000
}