Genetic programming for subjective fitness function identification

Dan Costelloe; Conor Ryan

doi:10.1007/978-3-540-24650-3_24

Genetic programming for subjective fitness function identification

Dan Costelloe, Conor Ryan

Department of Computer Science & Information Systems

University of Limerick

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

Abstract

This work addresses the common problem of modeling fitness functions for Interactive Evolutionary Systems. Such systems are necessarily slow because they need human interaction for the fundamental task of fitness allocation. The research presented here demonstrates that Genetic Programming can be used to learn subjective fitness functions from human subjects, using historical data from an Interactive Evolutionary system for producing pleasing drum patterns. The results indicate that GP is capable of performing symbolic regression even when the number of training cases is substantially less than the number of inputs.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Editors	Maarten Keijzer, Simon M. Lucas, Ernesto Costa, Terence Soule, Una-May O’Reilly
Publisher	Springer Verlag
Pages	259-268
Number of pages	10
ISBN (Print)	3540213465, 9783540213468
DOIs	https://doi.org/10.1007/978-3-540-24650-3_24
Publication status	Published - 2004

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3003
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Access to Document

10.1007/978-3-540-24650-3_24

Cite this

Costelloe, D., & Ryan, C. (2004). Genetic programming for subjective fitness function identification. In M. Keijzer, S. M. Lucas, E. Costa, T. Soule, & U.-M. O’Reilly (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 259-268). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3003). Springer Verlag. https://doi.org/10.1007/978-3-540-24650-3_24

Costelloe, Dan ; Ryan, Conor. / Genetic programming for subjective fitness function identification. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). editor / Maarten Keijzer ; Simon M. Lucas ; Ernesto Costa ; Terence Soule ; Una-May O’Reilly. Springer Verlag, 2004. pp. 259-268 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{f91ca7cd7e9245879ff67d522135c1bb,

title = "Genetic programming for subjective fitness function identification",

abstract = "This work addresses the common problem of modeling fitness functions for Interactive Evolutionary Systems. Such systems are necessarily slow because they need human interaction for the fundamental task of fitness allocation. The research presented here demonstrates that Genetic Programming can be used to learn subjective fitness functions from human subjects, using historical data from an Interactive Evolutionary system for producing pleasing drum patterns. The results indicate that GP is capable of performing symbolic regression even when the number of training cases is substantially less than the number of inputs.",

author = "Dan Costelloe and Conor Ryan",

year = "2004",

doi = "10.1007/978-3-540-24650-3_24",

language = "English",

isbn = "3540213465",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "259--268",

editor = "Maarten Keijzer and Lucas, {Simon M.} and Ernesto Costa and Terence Soule and Una-May O{\textquoteright}Reilly",

booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

Costelloe, D & Ryan, C 2004, Genetic programming for subjective fitness function identification. in M Keijzer, SM Lucas, E Costa, T Soule & U-M O’Reilly (eds), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3003, Springer Verlag, pp. 259-268. https://doi.org/10.1007/978-3-540-24650-3_24

Genetic programming for subjective fitness function identification. / Costelloe, Dan; Ryan, Conor.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). ed. / Maarten Keijzer; Simon M. Lucas; Ernesto Costa; Terence Soule; Una-May O’Reilly. Springer Verlag, 2004. p. 259-268 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3003).

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

TY - CHAP

T1 - Genetic programming for subjective fitness function identification

AU - Costelloe, Dan

AU - Ryan, Conor

PY - 2004

Y1 - 2004

N2 - This work addresses the common problem of modeling fitness functions for Interactive Evolutionary Systems. Such systems are necessarily slow because they need human interaction for the fundamental task of fitness allocation. The research presented here demonstrates that Genetic Programming can be used to learn subjective fitness functions from human subjects, using historical data from an Interactive Evolutionary system for producing pleasing drum patterns. The results indicate that GP is capable of performing symbolic regression even when the number of training cases is substantially less than the number of inputs.

AB - This work addresses the common problem of modeling fitness functions for Interactive Evolutionary Systems. Such systems are necessarily slow because they need human interaction for the fundamental task of fitness allocation. The research presented here demonstrates that Genetic Programming can be used to learn subjective fitness functions from human subjects, using historical data from an Interactive Evolutionary system for producing pleasing drum patterns. The results indicate that GP is capable of performing symbolic regression even when the number of training cases is substantially less than the number of inputs.

UR - http://www.scopus.com/inward/record.url?scp=34548102777&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-24650-3_24

DO - 10.1007/978-3-540-24650-3_24

M3 - Chapter

AN - SCOPUS:34548102777

SN - 3540213465

SN - 9783540213468

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 259

EP - 268

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

A2 - Keijzer, Maarten

A2 - Lucas, Simon M.

A2 - Costa, Ernesto

A2 - Soule, Terence

A2 - O’Reilly, Una-May

PB - Springer Verlag

ER -

Costelloe D, Ryan C. Genetic programming for subjective fitness function identification. In Keijzer M, Lucas SM, Costa E, Soule T, O’Reilly UM, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer Verlag. 2004. p. 259-268. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-24650-3_24

Genetic programming for subjective fitness function identification

Abstract

Publication series

Access to Document

Other files and links

Fingerprint

Cite this