Confidence intervals for computational effort comparisons

Matthew Walker; Howard Edwards; Chris Messom

Confidence intervals for computational effort comparisons

Matthew Walker, Howard Edwards, Chris Messom

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

Abstract

When researchers make alterations to the genetic programming algorithm they almost invariably wish to measure the change in performance of the evolutionary system. No one specific measure is standard, but Koza's computational effort statistic is frequently used [8]. In this paper the use of Koza's statistic is discussed and a study is made of three methods that produce confidence intervals for the statistic. It is found that an approximate 95% confidence interval can be easily produced.

Original language	English
Title of host publication	Genetic Programming - 10th European Conference, EuroGP 2007 Proceedings
Pages	23-32
Number of pages	10
Volume	4445 LNCS
Publication status	Published - 2007
Externally published	Yes
Event	10th European Conference on Genetic Programming, EuroGP 2007 - Valencia, Spain Duration: 11 Apr 2007 → 13 Apr 2007

Conference

Conference	10th European Conference on Genetic Programming, EuroGP 2007
Country/Territory	Spain
City	Valencia
Period	11/04/07 → 13/04/07

Cite this

@inproceedings{7b7b3373f5694e6d9ee02a9b2a6251c3,

title = "Confidence intervals for computational effort comparisons",

abstract = "When researchers make alterations to the genetic programming algorithm they almost invariably wish to measure the change in performance of the evolutionary system. No one specific measure is standard, but Koza's computational effort statistic is frequently used [8]. In this paper the use of Koza's statistic is discussed and a study is made of three methods that produce confidence intervals for the statistic. It is found that an approximate 95% confidence interval can be easily produced.",

author = "Matthew Walker and Howard Edwards and Chris Messom",

year = "2007",

language = "English",

isbn = "3540716025",

volume = "4445 LNCS",

pages = "23--32",

booktitle = "Genetic Programming - 10th European Conference, EuroGP 2007 Proceedings",

note = "10th European Conference on Genetic Programming, EuroGP 2007 ; Conference date: 11-04-2007 Through 13-04-2007",

}

TY - GEN

T1 - Confidence intervals for computational effort comparisons

AU - Walker, Matthew

AU - Edwards, Howard

AU - Messom, Chris

PY - 2007

Y1 - 2007

N2 - When researchers make alterations to the genetic programming algorithm they almost invariably wish to measure the change in performance of the evolutionary system. No one specific measure is standard, but Koza's computational effort statistic is frequently used [8]. In this paper the use of Koza's statistic is discussed and a study is made of three methods that produce confidence intervals for the statistic. It is found that an approximate 95% confidence interval can be easily produced.

AB - When researchers make alterations to the genetic programming algorithm they almost invariably wish to measure the change in performance of the evolutionary system. No one specific measure is standard, but Koza's computational effort statistic is frequently used [8]. In this paper the use of Koza's statistic is discussed and a study is made of three methods that produce confidence intervals for the statistic. It is found that an approximate 95% confidence interval can be easily produced.

UR - http://www.scopus.com/inward/record.url?scp=34548091374&partnerID=8YFLogxK

M3 - Conference Paper

AN - SCOPUS:34548091374

SN - 3540716025

SN - 9783540716020

VL - 4445 LNCS

SP - 23

EP - 32

BT - Genetic Programming - 10th European Conference, EuroGP 2007 Proceedings

T2 - 10th European Conference on Genetic Programming, EuroGP 2007

Y2 - 11 April 2007 through 13 April 2007

ER -

Confidence intervals for computational effort comparisons

Abstract

Conference

Other files and links

Cite this