Uniformity assessment for evolutionary multi-objective optimization

Miqing Li; Jinhua Zheng; Guixia Xiao

doi:10.1109/CEC.2008.4630861

Uniformity assessment for evolutionary multi-objective optimization

Miqing Li^*, Jinhua Zheng, Guixia Xiao

^*Corresponding author for this work

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

23 Citations (Scopus)

Abstract

Uniformity assessment of approximations of the Pareto-optimal set is an important issue in comparing the performance of multi-objective evolutionary algorithms. Although a number of performance metrics existed, many are applicable to low objective problems (2-3 objectives). In addition, most of the existed metrics are only applied to the final non-dominated set. In this paper, we suggest a running metric which evaluates the uniformity of solutions at every generation of a MOEA run. In particular, this metric can compare the uniformity of population with different size in any number of objectives. With an agglomeration of generation-wise populations, the metric reveals the change of uniformity in a MOEA run or helps provide a comparative evaluation of two or more MOEAs.

Original language	English
Title of host publication	2008 IEEE Congress on Evolutionary Computation, CEC 2008
Pages	625-632
Number of pages	8
DOIs	https://doi.org/10.1109/CEC.2008.4630861
Publication status	Published - 2008
Event	2008 IEEE Congress on Evolutionary Computation, CEC 2008 - Hong Kong, China Duration: 1 Jun 2008 → 6 Jun 2008

Publication series

Name	2008 IEEE Congress on Evolutionary Computation, CEC 2008

Conference

Conference	2008 IEEE Congress on Evolutionary Computation, CEC 2008
Country/Territory	China
City	Hong Kong
Period	1/06/08 → 6/06/08

ASJC Scopus subject areas

Computational Theory and Mathematics
Theoretical Computer Science

Access to Document

10.1109/CEC.2008.4630861

Cite this

@inproceedings{663b41b64a034548a61c3641368f247a,

title = "Uniformity assessment for evolutionary multi-objective optimization",

abstract = "Uniformity assessment of approximations of the Pareto-optimal set is an important issue in comparing the performance of multi-objective evolutionary algorithms. Although a number of performance metrics existed, many are applicable to low objective problems (2-3 objectives). In addition, most of the existed metrics are only applied to the final non-dominated set. In this paper, we suggest a running metric which evaluates the uniformity of solutions at every generation of a MOEA run. In particular, this metric can compare the uniformity of population with different size in any number of objectives. With an agglomeration of generation-wise populations, the metric reveals the change of uniformity in a MOEA run or helps provide a comparative evaluation of two or more MOEAs.",

author = "Miqing Li and Jinhua Zheng and Guixia Xiao",

year = "2008",

doi = "10.1109/CEC.2008.4630861",

language = "English",

isbn = "9781424418237",

series = "2008 IEEE Congress on Evolutionary Computation, CEC 2008",

pages = "625--632",

booktitle = "2008 IEEE Congress on Evolutionary Computation, CEC 2008",

note = "2008 IEEE Congress on Evolutionary Computation, CEC 2008 ; Conference date: 01-06-2008 Through 06-06-2008",

}

TY - GEN

T1 - Uniformity assessment for evolutionary multi-objective optimization

AU - Li, Miqing

AU - Zheng, Jinhua

AU - Xiao, Guixia

PY - 2008

Y1 - 2008

N2 - Uniformity assessment of approximations of the Pareto-optimal set is an important issue in comparing the performance of multi-objective evolutionary algorithms. Although a number of performance metrics existed, many are applicable to low objective problems (2-3 objectives). In addition, most of the existed metrics are only applied to the final non-dominated set. In this paper, we suggest a running metric which evaluates the uniformity of solutions at every generation of a MOEA run. In particular, this metric can compare the uniformity of population with different size in any number of objectives. With an agglomeration of generation-wise populations, the metric reveals the change of uniformity in a MOEA run or helps provide a comparative evaluation of two or more MOEAs.

AB - Uniformity assessment of approximations of the Pareto-optimal set is an important issue in comparing the performance of multi-objective evolutionary algorithms. Although a number of performance metrics existed, many are applicable to low objective problems (2-3 objectives). In addition, most of the existed metrics are only applied to the final non-dominated set. In this paper, we suggest a running metric which evaluates the uniformity of solutions at every generation of a MOEA run. In particular, this metric can compare the uniformity of population with different size in any number of objectives. With an agglomeration of generation-wise populations, the metric reveals the change of uniformity in a MOEA run or helps provide a comparative evaluation of two or more MOEAs.

UR - http://www.scopus.com/inward/record.url?scp=55749092280&partnerID=8YFLogxK

U2 - 10.1109/CEC.2008.4630861

DO - 10.1109/CEC.2008.4630861

M3 - Conference contribution

AN - SCOPUS:55749092280

SN - 9781424418237

T3 - 2008 IEEE Congress on Evolutionary Computation, CEC 2008

SP - 625

EP - 632

BT - 2008 IEEE Congress on Evolutionary Computation, CEC 2008

T2 - 2008 IEEE Congress on Evolutionary Computation, CEC 2008

Y2 - 1 June 2008 through 6 June 2008

ER -

Uniformity assessment for evolutionary multi-objective optimization

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this