Reference point specification in hypervolume calculation for fair comparison and efficient search

Hypervolume has been frequently used as a performance indicator for comparing evolutionary multiobjective optimization (EMO) algorithms. Hypervolume has been also used in indicator-based algorithms. Whereas a reference point is needed for hypervolume calculation, its specification has not been discussed in detail from a viewpoint of fair comparison. This may be because a slightly worse reference point than the nadir point seems to work well. In this paper, we tackle this issue: How to specify a reference point for fair comparison. First we discuss an appropriate specification of a reference point for multiobjective problems. Our discussions are based on the well-known theoretical results about the optimal solution distribution for hypervolume maximization. Next we examine various specifications by computational experiments. Experimental results show that a slightly worse reference point than the nadir point works well only for test problems with triangular Pareto fronts. Then we explain why this specification is not always appropriate for test problems with inverted triangular Pareto fronts. We also report a number of solution sets obtained by SMS-EMOA with various specifications of a reference point.

[1]  Qingfu Zhang,et al.  An Evolutionary Many-Objective Optimization Algorithm Based on Dominance and Decomposition , 2015, IEEE Transactions on Evolutionary Computation.

[2]  Qingfu Zhang,et al.  MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition , 2007, IEEE Transactions on Evolutionary Computation.

[3]  Bilel Derbel,et al.  Experiments on Greedy and Local Search Heuristics for ddimensional Hypervolume Subset Selection , 2016, GECCO.

[4]  Karl Bringmann,et al.  Two-dimensional subset selection for hypervolume and epsilon-indicator , 2014, GECCO.

[5]  Hisao Ishibuchi,et al.  Performance of Decomposition-Based Many-Objective Algorithms Strongly Depends on Pareto Front Shapes , 2017, IEEE Transactions on Evolutionary Computation.

[6]  Bo Zhang,et al.  Balancing Convergence and Diversity in Decomposition-Based Many-Objective Optimizers , 2016, IEEE Transactions on Evolutionary Computation.

[7]  Eckart Zitzler,et al.  HypE: An Algorithm for Fast Hypervolume-Based Many-Objective Optimization , 2011, Evolutionary Computation.

[8]  Kalyanmoy Deb,et al.  A Unified Evolutionary Optimization Procedure for Single, Multiple, and Many Objectives , 2016, IEEE Transactions on Evolutionary Computation.

[9]  Hisao Ishibuchi,et al.  Selecting a small number of non-dominated solutions to be presented to the decision maker , 2014, 2014 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[10]  Frank Neumann,et al.  Multiplicative Approximations, Optimal Hypervolume Distributions, and the Choice of the Reference Point , 2013, Evolutionary Computation.

[11]  Hisao Ishibuchi,et al.  Behavior of Multiobjective Evolutionary Algorithms on Many-Objective Knapsack Problems , 2015, IEEE Transactions on Evolutionary Computation.

[12]  Carlos M. Fonseca,et al.  Greedy Hypervolume Subset Selection in Low Dimensions , 2016, Evolutionary Computation.

[13]  Dimo Brockhoff,et al.  Optimal µ-Distributions for the Hypervolume Indicator for Problems with Linear Bi-objective Fronts: Exact and Exhaustive Results , 2010, SEAL.

[14]  Xin Yao,et al.  A New Dominance Relation-Based Evolutionary Algorithm for Many-Objective Optimization , 2016, IEEE Transactions on Evolutionary Computation.

[15]  Nicola Beume,et al.  SMS-EMOA: Multiobjective selection based on dominated hypervolume , 2007, Eur. J. Oper. Res..

[16]  Anne Auger,et al.  Hypervolume-based multiobjective optimization: Theoretical foundations and practical implications , 2012, Theor. Comput. Sci..

[17]  Marco Laumanns,et al.  Scalable multi-objective optimization test problems , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[18]  R. Lyndon While,et al.  A review of multiobjective test problems and a scalable test problem toolkit , 2006, IEEE Transactions on Evolutionary Computation.

[19]  Marco Laumanns,et al.  Performance assessment of multiobjective optimizers: an analysis and review , 2003, IEEE Trans. Evol. Comput..

[20]  Lothar Thiele,et al.  The Hypervolume Indicator Revisited: On the Design of Pareto-compliant Indicators Via Weighted Integration , 2007, EMO.

[21]  Lothar Thiele,et al.  Multiobjective Optimization Using Evolutionary Algorithms - A Comparative Case Study , 1998, PPSN.

[22]  Carlos M. Fonseca,et al.  Hypervolume Subset Selection in Two Dimensions: Formulations and Algorithms , 2016, Evolutionary Computation.

[23]  Hisao Ishibuchi,et al.  Hypervolume Subset Selection for Triangular and Inverted Triangular Pareto Fronts of Three-Objective Problems , 2017, FOGA '17.

[24]  Carlos M. Fonseca,et al.  Greedy Hypervolume Subset Selection in the Three-Objective Case , 2015, GECCO.

[25]  Nicola Beume,et al.  Pareto-, Aggregation-, and Indicator-Based Methods in Many-Objective Optimization , 2007, EMO.

[26]  Hartmut Schmeck,et al.  A theoretical analysis of volume based Pareto front approximations , 2014, GECCO.

[27]  Kalyanmoy Deb,et al.  An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point Based Nondominated Sorting Approach, Part II: Handling Constraints and Extending to an Adaptive Approach , 2014, IEEE Transactions on Evolutionary Computation.