An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods

Miqing LI; Xin YAO

doi:10.1007/978-3-030-12598-1_2

An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods

Miqing LI^*, Xin YAO

^*Corresponding author for this work

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

41 Citations (SciVal)

Abstract

Most evolutionary multiobjective optimisation (EMO) algorithms explicitly or implicitly maintain an archive for an approximation of the Pareto front. A question arising is whether existing archiving methods are reliable with respect to their convergence and approximation ability. Despite theoretical results available, it remains unknown how these archivers actually perform in practice. In particular, what percentage of solutions in their final archive are Pareto optimal? How frequently do they experience deterioration during the archiving process? Deterioration means archiving a new solution which is dominated by some solution discarded previously. This paper answers the above questions through a systematic investigation of eight representative archivers on 37 test instances with two to five objectives. We have found that (1) deterioration happens to all the archivers; (2) the deterioration degree can vary dramatically on different problems; (3) some archivers clearly perform better than others; and (4) several popular archivers sometime return a population with most solutions being the non-optimal. All of these suggest the need of improvement of current archiving methods. © Springer Nature Switzerland AG 2019.

Original language	English
Title of host publication	Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings
Editors	Kalyanmoy DEB, Erik GOODMAN, Carlos A. Coello COELLO, Kathrin KLAMROTH, Kaisa MIETTINEN, Sanaz MOSTAGHIM, Patrick REED
Publisher	Springer
Pages	15-26
Number of pages	12
ISBN (Electronic)	9783030125981
ISBN (Print)	9783030125974
DOIs	https://doi.org/10.1007/978-3-030-12598-1_2
Publication status	Published - 2019
Externally published	Yes
Event	10th International Conference on Evolutionary Multi-Criterion Optimization - East Lansing, United States Duration: 10 Mar 2019 → 13 Mar 2019

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	11411
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	10th International Conference on Evolutionary Multi-Criterion Optimization
Abbreviated title	EMO 2019
Country/Territory	United States
City	East Lansing
Period	10/03/19 → 13/03/19

Bibliographical note

Publisher Copyright:
© Springer Nature Switzerland AG 2019.

Funding

This work was supported by EPSRC (Grant Nos. EP/J017515/1 and EP/P005578/1) and Science and Technology Innovation Committee Foundation of Shenzhen (Grant Nos. ZDSYS201703031748284 and JCYJ20170307105521943).

Keywords

Archive
Empirical investigation
Evolutionary computation
Monotonicity
Multi-objective optimisation
Optimality

Access to Document

10.1007/978-3-030-12598-1_2

Cite this

LI, M., & YAO, X. (2019). An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods. In K. DEB, E. GOODMAN, C. A. C. COELLO, K. KLAMROTH, K. MIETTINEN, S. MOSTAGHIM, & P. REED (Eds.), Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings (pp. 15-26). (Lecture Notes in Computer Science; Vol. 11411). Springer. https://doi.org/10.1007/978-3-030-12598-1_2

LI, Miqing ; YAO, Xin. / An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods. Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings. editor / Kalyanmoy DEB ; Erik GOODMAN ; Carlos A. Coello COELLO ; Kathrin KLAMROTH ; Kaisa MIETTINEN ; Sanaz MOSTAGHIM ; Patrick REED. Springer, 2019. pp. 15-26 (Lecture Notes in Computer Science).

@inproceedings{8f83b12add9249f68f7cf7dcf118207f,

title = "An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods",

abstract = "Most evolutionary multiobjective optimisation (EMO) algorithms explicitly or implicitly maintain an archive for an approximation of the Pareto front. A question arising is whether existing archiving methods are reliable with respect to their convergence and approximation ability. Despite theoretical results available, it remains unknown how these archivers actually perform in practice. In particular, what percentage of solutions in their final archive are Pareto optimal? How frequently do they experience deterioration during the archiving process? Deterioration means archiving a new solution which is dominated by some solution discarded previously. This paper answers the above questions through a systematic investigation of eight representative archivers on 37 test instances with two to five objectives. We have found that (1) deterioration happens to all the archivers; (2) the deterioration degree can vary dramatically on different problems; (3) some archivers clearly perform better than others; and (4) several popular archivers sometime return a population with most solutions being the non-optimal. All of these suggest the need of improvement of current archiving methods. {\textcopyright} Springer Nature Switzerland AG 2019.",

keywords = "Archive, Empirical investigation, Evolutionary computation, Monotonicity, Multi-objective optimisation, Optimality",

author = "Miqing LI and Xin YAO",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.; 10th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2019 ; Conference date: 10-03-2019 Through 13-03-2019",

year = "2019",

doi = "10.1007/978-3-030-12598-1_2",

language = "English",

isbn = "9783030125974",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "15--26",

editor = "DEB, {Kalyanmoy } and GOODMAN, {Erik } and COELLO, {Carlos A. Coello } and KLAMROTH, {Kathrin } and MIETTINEN, {Kaisa } and MOSTAGHIM, {Sanaz } and REED, {Patrick }",

booktitle = "Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings",

}

LI, M & YAO, X 2019, An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods. in K DEB, E GOODMAN, CAC COELLO, K KLAMROTH, K MIETTINEN, S MOSTAGHIM & P REED (eds), Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings. Lecture Notes in Computer Science, vol. 11411, Springer, pp. 15-26, 10th International Conference on Evolutionary Multi-Criterion Optimization, East Lansing, Michigan, United States, 10/03/19. https://doi.org/10.1007/978-3-030-12598-1_2

An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods. / LI, Miqing; YAO, Xin.
Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings. ed. / Kalyanmoy DEB; Erik GOODMAN; Carlos A. Coello COELLO; Kathrin KLAMROTH; Kaisa MIETTINEN; Sanaz MOSTAGHIM; Patrick REED. Springer, 2019. p. 15-26 (Lecture Notes in Computer Science; Vol. 11411).

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

TY - GEN

T1 - An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods

AU - LI, Miqing

AU - YAO, Xin

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2019.

PY - 2019

Y1 - 2019

N2 - Most evolutionary multiobjective optimisation (EMO) algorithms explicitly or implicitly maintain an archive for an approximation of the Pareto front. A question arising is whether existing archiving methods are reliable with respect to their convergence and approximation ability. Despite theoretical results available, it remains unknown how these archivers actually perform in practice. In particular, what percentage of solutions in their final archive are Pareto optimal? How frequently do they experience deterioration during the archiving process? Deterioration means archiving a new solution which is dominated by some solution discarded previously. This paper answers the above questions through a systematic investigation of eight representative archivers on 37 test instances with two to five objectives. We have found that (1) deterioration happens to all the archivers; (2) the deterioration degree can vary dramatically on different problems; (3) some archivers clearly perform better than others; and (4) several popular archivers sometime return a population with most solutions being the non-optimal. All of these suggest the need of improvement of current archiving methods. © Springer Nature Switzerland AG 2019.

AB - Most evolutionary multiobjective optimisation (EMO) algorithms explicitly or implicitly maintain an archive for an approximation of the Pareto front. A question arising is whether existing archiving methods are reliable with respect to their convergence and approximation ability. Despite theoretical results available, it remains unknown how these archivers actually perform in practice. In particular, what percentage of solutions in their final archive are Pareto optimal? How frequently do they experience deterioration during the archiving process? Deterioration means archiving a new solution which is dominated by some solution discarded previously. This paper answers the above questions through a systematic investigation of eight representative archivers on 37 test instances with two to five objectives. We have found that (1) deterioration happens to all the archivers; (2) the deterioration degree can vary dramatically on different problems; (3) some archivers clearly perform better than others; and (4) several popular archivers sometime return a population with most solutions being the non-optimal. All of these suggest the need of improvement of current archiving methods. © Springer Nature Switzerland AG 2019.

KW - Archive

KW - Empirical investigation

KW - Evolutionary computation

KW - Monotonicity

KW - Multi-objective optimisation

KW - Optimality

UR - http://www.scopus.com/inward/record.url?scp=85063061758&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-12598-1_2

DO - 10.1007/978-3-030-12598-1_2

M3 - Conference paper (refereed)

SN - 9783030125974

T3 - Lecture Notes in Computer Science

SP - 15

EP - 26

BT - Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings

A2 - DEB, Kalyanmoy

A2 - GOODMAN, Erik

A2 - COELLO, Carlos A. Coello

A2 - KLAMROTH, Kathrin

A2 - MIETTINEN, Kaisa

A2 - MOSTAGHIM, Sanaz

A2 - REED, Patrick

PB - Springer

T2 - 10th International Conference on Evolutionary Multi-Criterion Optimization

Y2 - 10 March 2019 through 13 March 2019

ER -

LI M, YAO X. An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods. In DEB K, GOODMAN E, COELLO CAC, KLAMROTH K, MIETTINEN K, MOSTAGHIM S, REED P, editors, Evolutionary Multi-Criterion Optimization : 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings. Springer. 2019. p. 15-26. (Lecture Notes in Computer Science). Epub 2019 Feb 3. doi: 10.1007/978-3-030-12598-1_2

An empirical investigation of the optimality and monotonicity properties of multiobjective archiving methods

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this