Convergence versus diversity in multiobjective optimization

Shouyong Jiang; Shengxiang Yang

doi:10.1007/978-3-319-45823-6_92

Convergence versus diversity in multiobjective optimization

Shouyong Jiang^*, Shengxiang Yang

^*Corresponding author for this work

Computing Science

De Montfort University

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

14 Citations (Scopus)

Abstract

Convergence and diversity are two main goals in multiobjective optimization. In literature, most existing multiobjective optimization evolutionary algorithms (MOEAs) adopt a convergencefirst- and-diversity-second environmental selection which prefers nondominated solutions to dominated ones, as is the case with the popular nondominated sorting based selection method. While convergence-first sorting has continuously shown effectiveness for handling a variety of problems, it faces challenges to maintain well population diversity due to the overemphasis of convergence. In this paper, we propose a general diversity-first sorting method for multiobjective optimization. Based on the method, a new MOEA, called DBEA, is then introduced. DBEA is compared with the recently-developed nondominated sorting genetic algorithm III (NSGA-III) on different problems. Experimental studies show that the diversity-first method has great potential for diversity maintenance and is very competitive for many-objective optimization.

Original language	English
Title of host publication	Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings
Editors	Emma Hart, Ben Paechter, Julia Handl, Manuel López-Ibáñez, Peter R. Lewis, Gabriela Ochoa
Publisher	Springer Verlag
Pages	984-993
Number of pages	10
ISBN (Print)	9783319458229
DOIs	https://doi.org/10.1007/978-3-319-45823-6_92
Publication status	Published - 2016
Event	14th International Conference on Parallel Problem Solving from Nature, PPSN 2016 - Edinburgh, United Kingdom Duration: 17 Sept 2016 → 21 Sept 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9921 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th International Conference on Parallel Problem Solving from Nature, PPSN 2016
Country/Territory	United Kingdom
City	Edinburgh
Period	17/09/16 → 21/09/16

Bibliographical note

Funding Information:
This work was funded by the Engineering and Physical Sciences Research Council (EPSRC) of U.K. under Grant EP/K001310/1.

Publisher Copyright:
© Springer International Publishing AG 2016.

Access to Document

10.1007/978-3-319-45823-6_92

Cite this

Jiang, S., & Yang, S. (2016). Convergence versus diversity in multiobjective optimization. In E. Hart, B. Paechter, J. Handl, M. López-Ibáñez, P. R. Lewis, & G. Ochoa (Eds.), Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings (pp. 984-993). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9921 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-45823-6_92

Convergence versus diversity in multiobjective optimization. / Jiang, Shouyong; Yang, Shengxiang.
Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings. ed. / Emma Hart; Ben Paechter; Julia Handl; Manuel López-Ibáñez; Peter R. Lewis; Gabriela Ochoa. Springer Verlag, 2016. p. 984-993 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9921 LNCS).

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

Jiang, S & Yang, S 2016, Convergence versus diversity in multiobjective optimization. in E Hart, B Paechter, J Handl, M López-Ibáñez, PR Lewis & G Ochoa (eds), Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9921 LNCS, Springer Verlag, pp. 984-993, 14th International Conference on Parallel Problem Solving from Nature, PPSN 2016, Edinburgh, United Kingdom, 17/09/16. https://doi.org/10.1007/978-3-319-45823-6_92

Jiang S, Yang S. Convergence versus diversity in multiobjective optimization. In Hart E, Paechter B, Handl J, López-Ibáñez M, Lewis PR, Ochoa G, editors, Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings. Springer Verlag. 2016. p. 984-993. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Epub 2016 Aug 31. doi: 10.1007/978-3-319-45823-6_92

Jiang, Shouyong ; Yang, Shengxiang. / Convergence versus diversity in multiobjective optimization. Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings. editor / Emma Hart ; Ben Paechter ; Julia Handl ; Manuel López-Ibáñez ; Peter R. Lewis ; Gabriela Ochoa. Springer Verlag, 2016. pp. 984-993 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{6bfdd1a4abe84eb6889897d5ce1581c7,

title = "Convergence versus diversity in multiobjective optimization",

abstract = "Convergence and diversity are two main goals in multiobjective optimization. In literature, most existing multiobjective optimization evolutionary algorithms (MOEAs) adopt a convergencefirst- and-diversity-second environmental selection which prefers nondominated solutions to dominated ones, as is the case with the popular nondominated sorting based selection method. While convergence-first sorting has continuously shown effectiveness for handling a variety of problems, it faces challenges to maintain well population diversity due to the overemphasis of convergence. In this paper, we propose a general diversity-first sorting method for multiobjective optimization. Based on the method, a new MOEA, called DBEA, is then introduced. DBEA is compared with the recently-developed nondominated sorting genetic algorithm III (NSGA-III) on different problems. Experimental studies show that the diversity-first method has great potential for diversity maintenance and is very competitive for many-objective optimization.",

author = "Shouyong Jiang and Shengxiang Yang",

note = "Funding Information: This work was funded by the Engineering and Physical Sciences Research Council (EPSRC) of U.K. under Grant EP/K001310/1. Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 14th International Conference on Parallel Problem Solving from Nature, PPSN 2016 ; Conference date: 17-09-2016 Through 21-09-2016",

year = "2016",

doi = "10.1007/978-3-319-45823-6_92",

language = "English",

isbn = "9783319458229",

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

publisher = "Springer Verlag",

pages = "984--993",

editor = "Emma Hart and Ben Paechter and Julia Handl and Manuel L{\'o}pez-Ib{\'a}{\~n}ez and Lewis, {Peter R.} and Gabriela Ochoa",

booktitle = "Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings",

address = "Germany",

}

TY - GEN

T1 - Convergence versus diversity in multiobjective optimization

AU - Jiang, Shouyong

AU - Yang, Shengxiang

N1 - Funding Information: This work was funded by the Engineering and Physical Sciences Research Council (EPSRC) of U.K. under Grant EP/K001310/1. Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - Convergence and diversity are two main goals in multiobjective optimization. In literature, most existing multiobjective optimization evolutionary algorithms (MOEAs) adopt a convergencefirst- and-diversity-second environmental selection which prefers nondominated solutions to dominated ones, as is the case with the popular nondominated sorting based selection method. While convergence-first sorting has continuously shown effectiveness for handling a variety of problems, it faces challenges to maintain well population diversity due to the overemphasis of convergence. In this paper, we propose a general diversity-first sorting method for multiobjective optimization. Based on the method, a new MOEA, called DBEA, is then introduced. DBEA is compared with the recently-developed nondominated sorting genetic algorithm III (NSGA-III) on different problems. Experimental studies show that the diversity-first method has great potential for diversity maintenance and is very competitive for many-objective optimization.

AB - Convergence and diversity are two main goals in multiobjective optimization. In literature, most existing multiobjective optimization evolutionary algorithms (MOEAs) adopt a convergencefirst- and-diversity-second environmental selection which prefers nondominated solutions to dominated ones, as is the case with the popular nondominated sorting based selection method. While convergence-first sorting has continuously shown effectiveness for handling a variety of problems, it faces challenges to maintain well population diversity due to the overemphasis of convergence. In this paper, we propose a general diversity-first sorting method for multiobjective optimization. Based on the method, a new MOEA, called DBEA, is then introduced. DBEA is compared with the recently-developed nondominated sorting genetic algorithm III (NSGA-III) on different problems. Experimental studies show that the diversity-first method has great potential for diversity maintenance and is very competitive for many-objective optimization.

UR - http://www.scopus.com/inward/record.url?scp=84988499028&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-45823-6_92

DO - 10.1007/978-3-319-45823-6_92

M3 - Published conference contribution

AN - SCOPUS:84988499028

SN - 9783319458229

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

SP - 984

EP - 993

BT - Parallel Problem Solving from Nature - 14th International Conference, PPSN 2016, Proceedings

A2 - Hart, Emma

A2 - Paechter, Ben

A2 - Handl, Julia

A2 - López-Ibáñez, Manuel

A2 - Lewis, Peter R.

A2 - Ochoa, Gabriela

PB - Springer Verlag

T2 - 14th International Conference on Parallel Problem Solving from Nature, PPSN 2016

Y2 - 17 September 2016 through 21 September 2016

ER -

Convergence versus diversity in multiobjective optimization

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this