How to Specify a Reference Point in Hypervolume Calculation for Fair Performance Comparison

被引:166
作者
Ishibuchi, Hisao [1 ]
Imada, Ryo [2 ]
Setoguchi, Yu [2 ]
Nojima, Yusuke [2 ]
机构
[1] Southern Univ Sci & Technol, Shenzhen Key Lab Computat Intelligence, Dept Comp Sci & Engn, Shenzhen 518005, Peoples R China
[2] Osaka Prefecture Univ, Dept Comp Sci & Intelligent Syst, Sakai, Osaka 5998531, Japan
关键词
Evolutionary multi-objective optimization; hypervolume; reference point; performance comparison; EVOLUTIONARY ALGORITHMS; PARETO; APPROXIMATIONS; INDICATORS; SELECTION;
D O I
10.1162/evco_a_00226
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The hypervolume indicator has frequently been used for comparing evolutionary multi-objective optimization (EMO) algorithms. A reference point is needed for hypervolume calculation. However, its specification has not been discussed in detail from a viewpoint of fair performance comparison. A slightly worse point than the nadir point is usually used for hypervolume calculation in the EMO community. In this paper, we propose a reference point specification method for fair performance comparison of EMO algorithms. First, we discuss the relation between the reference point specification and the optimal distribution of solutions for hypervolume maximization. It is demonstrated that the optimal distribution of solutions strongly depends on the location of the reference point when a multi-objective problem has an inverted triangular Pareto front. Next, we propose a reference point specification method based on theoretical discussions on the optimal distribution of solutions. The basic idea is to specify the reference point so that a set of well-distributed solutions over the entire linear Pareto front has a large hypervolume and all solutions in such a solution set have similar hypervolume contributions. Then, we examine whether the proposed method can appropriately specify the reference point through computational experiments on various test problems. Finally, we examine the usefulness of the proposed method in a hypervolume-based EMO algorithm. Our discussions and experimental results clearly show that a slightly worse point than the nadir point is not always appropriate for performance comparison of EMO algorithms.
引用
收藏
页码:411 / 440
页数:30
相关论文
共 43 条
[1]  
[Anonymous], THESIS
[2]   A Decomposition-Based Evolutionary Algorithm for Many Objective Optimization [J].
Asafuddoula, M. ;
Ray, Tapabrata ;
Sarker, Ruhul .
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2015, 19 (03) :445-460
[3]   Hypervolume-based multiobjective optimization: Theoretical foundations and practical implications [J].
Auger, Anne ;
Bader, Johannes ;
Brockhoff, Dimo ;
Zitzler, Eckart .
THEORETICAL COMPUTER SCIENCE, 2012, 425 :75-103
[4]   HypE: An Algorithm for Fast Hypervolume-Based Many-Objective Optimization [J].
Bader, Johannes ;
Zitzler, Eckart .
EVOLUTIONARY COMPUTATION, 2011, 19 (01) :45-76
[5]   Experiments on Greedy and Local Search Heuristics for d-dimensional Hypervolume Subset Selection [J].
Basseur, Matthieu ;
Derbel, Bilel ;
Goeffon, Adrien ;
Liefooghe, Arnaud .
GECCO'16: PROCEEDINGS OF THE 2016 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2016, :541-548
[6]   SMS-EMOA: Multiobjective selection based on dominated hypervolume [J].
Beume, Nicola ;
Naujoks, Boris ;
Emmerich, Michael .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2007, 181 (03) :1653-1669
[7]   Two-dimensional Subset Selection for Hypervolume and Epsilon-Indicator [J].
Bringmann, Karl ;
Friedrich, Tobias ;
Klitzke, Patrick .
GECCO'14: PROCEEDINGS OF THE 2014 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2014, :589-596
[8]  
Brockhoff D, 2010, 8 INT C SIM EV LEARN, P24
[9]   A Surrogate-Assisted Reference Vector Guided Evolutionary Algorithm for Computationally Expensive Many-Objective Optimization [J].
Chugh, Tinkle ;
Jin, Yaochu ;
Miettinen, Kaisa ;
Hakanen, Jussi ;
Sindhya, Karthik .
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2018, 22 (01) :129-142
[10]  
Deb K, 2002, IEEE C EVOL COMPUTAT, P825, DOI 10.1109/CEC.2002.1007032