Improving the Tartarus problem as a benchmark in genetic programming

Thomas D. Griffiths*, Anikó Ekárt

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

For empirical research on computer algorithms, it is essential to have a set of benchmark problems on which the relative performance of different methods and their applicability can be assessed. In the majority of computational research fields there are established sets of benchmark problems; however, the field of genetic programming lacks a similarly rigorously defined set of benchmarks. There is a strong interest within the genetic programming community to develop a suite of benchmarks. Following recent surveys [7], the desirable characteristics of a benchmark problem are now better defined. In this paper the Tartarus problem is proposed as a tunably difficult benchmark problem for use in Genetic Programming. The justification for this proposal is presented, together with guidance on its usage as a benchmark.

Original languageEnglish
Title of host publicationGenetic programming
Subtitle of host publication20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings
EditorsJames McDermott, Mauro Castelli, Lukas Sekanina, et al
Place of PublicationCham (CH)
PublisherSpringer
Pages278-293
Number of pages16
ISBN (Electronic)978-3-319-55696-3
ISBN (Print)978-3-319-55695-6
DOIs
Publication statusE-pub ahead of print - 15 Mar 2017
Event20th European Conference on Genetic Programming, EuroGP 2017 - Amsterdam, Netherlands
Duration: 19 Apr 201721 Apr 2017

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume10196
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th European Conference on Genetic Programming, EuroGP 2017
CountryNetherlands
City Amsterdam
Period19/04/1721/04/17

Fingerprint

Genetic programming
Genetic Programming
Benchmark
Empirical Research
Justification
Guidance

Keywords

  • benchmark
  • genetic programming
  • tartarus

Cite this

Griffiths, T. D., & Ekárt, A. (2017). Improving the Tartarus problem as a benchmark in genetic programming. In J. McDermott, M. Castelli, L. Sekanina, & et al (Eds.), Genetic programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings (pp. 278-293). (Lecture Notes in Computer Science; Vol. 10196). Cham (CH): Springer. https://doi.org/10.1007/978-3-319-55696-3_18
Griffiths, Thomas D. ; Ekárt, Anikó. / Improving the Tartarus problem as a benchmark in genetic programming. Genetic programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings. editor / James McDermott ; Mauro Castelli ; Lukas Sekanina ; et al. Cham (CH) : Springer, 2017. pp. 278-293 (Lecture Notes in Computer Science).
@inproceedings{2db10058697f4259b388cc8d5c760e9b,
title = "Improving the Tartarus problem as a benchmark in genetic programming",
abstract = "For empirical research on computer algorithms, it is essential to have a set of benchmark problems on which the relative performance of different methods and their applicability can be assessed. In the majority of computational research fields there are established sets of benchmark problems; however, the field of genetic programming lacks a similarly rigorously defined set of benchmarks. There is a strong interest within the genetic programming community to develop a suite of benchmarks. Following recent surveys [7], the desirable characteristics of a benchmark problem are now better defined. In this paper the Tartarus problem is proposed as a tunably difficult benchmark problem for use in Genetic Programming. The justification for this proposal is presented, together with guidance on its usage as a benchmark.",
keywords = "benchmark, genetic programming, tartarus",
author = "Griffiths, {Thomas D.} and Anik{\'o} Ek{\'a}rt",
year = "2017",
month = "3",
day = "15",
doi = "10.1007/978-3-319-55696-3_18",
language = "English",
isbn = "978-3-319-55695-6",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "278--293",
editor = "James McDermott and Mauro Castelli and Lukas Sekanina and {et al}",
booktitle = "Genetic programming",
address = "Germany",

}

Griffiths, TD & Ekárt, A 2017, Improving the Tartarus problem as a benchmark in genetic programming. in J McDermott, M Castelli, L Sekanina & et al (eds), Genetic programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings. Lecture Notes in Computer Science, vol. 10196, Springer, Cham (CH), pp. 278-293, 20th European Conference on Genetic Programming, EuroGP 2017, Amsterdam, Netherlands, 19/04/17. https://doi.org/10.1007/978-3-319-55696-3_18

Improving the Tartarus problem as a benchmark in genetic programming. / Griffiths, Thomas D.; Ekárt, Anikó.

Genetic programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings. ed. / James McDermott; Mauro Castelli; Lukas Sekanina; et al. Cham (CH) : Springer, 2017. p. 278-293 (Lecture Notes in Computer Science; Vol. 10196).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Improving the Tartarus problem as a benchmark in genetic programming

AU - Griffiths, Thomas D.

AU - Ekárt, Anikó

PY - 2017/3/15

Y1 - 2017/3/15

N2 - For empirical research on computer algorithms, it is essential to have a set of benchmark problems on which the relative performance of different methods and their applicability can be assessed. In the majority of computational research fields there are established sets of benchmark problems; however, the field of genetic programming lacks a similarly rigorously defined set of benchmarks. There is a strong interest within the genetic programming community to develop a suite of benchmarks. Following recent surveys [7], the desirable characteristics of a benchmark problem are now better defined. In this paper the Tartarus problem is proposed as a tunably difficult benchmark problem for use in Genetic Programming. The justification for this proposal is presented, together with guidance on its usage as a benchmark.

AB - For empirical research on computer algorithms, it is essential to have a set of benchmark problems on which the relative performance of different methods and their applicability can be assessed. In the majority of computational research fields there are established sets of benchmark problems; however, the field of genetic programming lacks a similarly rigorously defined set of benchmarks. There is a strong interest within the genetic programming community to develop a suite of benchmarks. Following recent surveys [7], the desirable characteristics of a benchmark problem are now better defined. In this paper the Tartarus problem is proposed as a tunably difficult benchmark problem for use in Genetic Programming. The justification for this proposal is presented, together with guidance on its usage as a benchmark.

KW - benchmark

KW - genetic programming

KW - tartarus

UR - http://link.springer.com/chapter/10.1007%2F978-3-319-55696-3_18

UR - http://www.scopus.com/inward/record.url?scp=85015776887&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-55696-3_18

DO - 10.1007/978-3-319-55696-3_18

M3 - Conference contribution

AN - SCOPUS:85015776887

SN - 978-3-319-55695-6

T3 - Lecture Notes in Computer Science

SP - 278

EP - 293

BT - Genetic programming

A2 - McDermott, James

A2 - Castelli, Mauro

A2 - Sekanina, Lukas

A2 - et al,

PB - Springer

CY - Cham (CH)

ER -

Griffiths TD, Ekárt A. Improving the Tartarus problem as a benchmark in genetic programming. In McDermott J, Castelli M, Sekanina L, et al, editors, Genetic programming: 20th European Conference, EuroGP 2017, Amsterdam, The Netherlands, April 19-21, 2017, proceedings. Cham (CH): Springer. 2017. p. 278-293. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-319-55696-3_18