中文版 | English
Title

Analytical Methods to Separately Evaluate Convergence and Diversity for Multi-objective Optimization

Author
Corresponding AuthorMasuyama, Naoki
DOI
Publication Years
2023
Conference Name
14th Metaheuristics International Conference, MIC 2022
ISSN
0302-9743
EISSN
1611-3349
ISBN
9783031265037
Source Title
Volume
13838 LNCS
Pages
172-186
Conference Date
July 11, 2022 - July 14, 2022
Conference Place
Ortigia-Syracuse, Italy
Publisher
Abstract
This paper proposes two analytical methods which completely separate the search performance of multi-objective evolutionary algorithms (MOEAs) into convergence and diversity for quantitatively comparing MOEAs. Specifically, Convergence-Diversity Pair (C-D Pair) is proposed to statistically compare the convergence and diversity of two MOEAs. C-D Pair provides analytical information on the overall experimental results. In addition, Convergence-Diversity Diagram (C-D Diagram) is also proposed to visualize a pair of convergence and diversity of a solution set as a single point in a two-dimensional space. C-D Diagram enables a detailed and intuitive comparison of the search performance trends of multiple MOEAs. Moreover, this paper introduces two diversity indicators. These indicators are designed to evaluate only the diversity of the population in an MOEA by completely eliminating the effect of the convergence. Computational experiments demonstrate the analytical capability and validity of the proposed analytical methods by using various test problems.
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
SUSTech Authorship
Others
Language
English
Indexed By
EI Accession Number
20231113699303
EI Keywords
Evolutionary algorithms
ESI Classification Code
Optimization Techniques:921.5
Data Source
EV Compendex
Citation statistics
Cited Times [WOS]:0
Document TypeConference paper
Identifierhttp://kc.sustech.edu.cn/handle/2SGJ60CL/519744
DepartmentDepartment of Computer Science and Engineering
Affiliation
1.Graduate School of Engineering, Osaka Prefecture University, 1-1 Gakuen-cho Naka-ku, Osaka, Sakai-Shi; 599-8531, Japan
2.Graduate School of Informatics, Osaka Metropolitan University, 1-1 Gakuen-cho Naka-ku, Osaka, Sakai-Shi; 599-8531, Japan
3.Guangdong Provincial Key Laboratory of Brain-inspired Intelligent Computation, Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen; 518055, China
Recommended Citation
GB/T 7714
Kinoshita, Takato,Masuyama, Naoki,Nojima, Yusuke,et al. Analytical Methods to Separately Evaluate Convergence and Diversity for Multi-objective Optimization[C]:Springer Science and Business Media Deutschland GmbH,2023:172-186.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Export to Excel
Export to Csv
Altmetrics Score
Google Scholar
Similar articles in Google Scholar
[Kinoshita, Takato]'s Articles
[Masuyama, Naoki]'s Articles
[Nojima, Yusuke]'s Articles
Baidu Scholar
Similar articles in Baidu Scholar
[Kinoshita, Takato]'s Articles
[Masuyama, Naoki]'s Articles
[Nojima, Yusuke]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Kinoshita, Takato]'s Articles
[Masuyama, Naoki]'s Articles
[Nojima, Yusuke]'s Articles
Terms of Use
No data!
Social Bookmark/Share
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.