Abstract
Multiobjective evolutionary algorithms (MOEAs) have been widely used in real-world applications. However, most MOEAs based on Pareto-dominance handle many-objective problems (MaOPs) poorly due to a high proportion of incomparable and thus mutually nondominated solutions. Recently, a number ofmany-objective evolutionary algorithms (MaOEAs) have been proposed to deal with this scalability issue. In this article, a survey of MaOEAs is reported. According to the key ideas used, MaOEAs are categorized into seven classes: relaxed dominance based, diversity-based, aggregation-based, indicator-based, reference set based, preference-based, and dimensionality reduction approaches. Several future research directions in this field are also discussed.
Original language | English |
---|---|
Article number | A10 |
Journal | ACM Computing Surveys |
Volume | 48 |
Issue number | 1 |
DOIs | |
Publication status | Published - 29 Sept 2015 |
Externally published | Yes |
Funding
This work was supported in part by the 973 Program of China under grant 2011CB707006, the National Natural Science Foundation of China under grants 61175065 and 61329302, the Program for New Century Excellent Talents in University under grant NCET-12-0512, the Science and Technological Fund of Anhui Province for Outstanding Youth under grant 1108085J16, EPSRC grant EP/J017515/1, and the European Union Seventh Framework Programme under grant 247619. Xin Yao was supported by a Royal Society Wolfson Research Merit Award.
Keywords
- Evolutionary algorithm
- Many-objective optimization
- Scalability