Data-Driven Analysis of Pareto Set Topology

被引:4
作者
Hamada, Naoki [1 ]
Goto, Keisuke [1 ]
机构
[1] Fujitsu Labs Ltd, Kawasaki, Kanagawa, Japan
来源
GECCO'18: PROCEEDINGS OF THE 2018 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE | 2018年
关键词
multi-objective optimization; continuous optimization; topological data analysis; persistent homology;
D O I
10.1145/3205455.3205613
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
When and why can evolutionary multi-objective optimization (EMO) algorithms cover the entire Pareto set? That is a major concern for EMO researchers and practitioners. A recent theoretical study revealed that (roughly speaking) if the Pareto set forms a topological simplex (a curved line, a curved triangle, a curved tetrahedron, etc.), then decomposition-based EMO algorithms can cover the entire Pareto set. Usually, we cannot know the true Pareto set and have to estimate its topology by using the population of EMO algorithms during or after the runtime. This paper presents a data-driven approach to analyze the topology of the Pareto set. We give a theory of how to recognize the topology of the Pareto set from data and implement an algorithm to judge whether the true Pareto set may form a topological simplex or not. Numerical experiments show that the proposed method correctly recognizes the topology of high-dimensional Pareto sets within reasonable population size.
引用
收藏
页码:657 / 664
页数:8
相关论文
共 22 条
  • [1] Bendich P., 2010, ARXIV10083572
  • [2] Inferring local homology from sampled stratified spaces
    Bendich, Paul
    Cohen-Steiner, David
    Edelsbrunner, Herbert
    Harer, John
    Morozov, Dmitriy
    [J]. 48TH ANNUAL IEEE SYMPOSIUM ON FOUNDATIONS OF COMPUTER SCIENCE, PROCEEDINGS, 2007, : 536 - 546
  • [3] Persistent Intersection Homology
    Bendich, Paul
    Harer, John
    [J]. FOUNDATIONS OF COMPUTATIONAL MATHEMATICS, 2011, 11 (03) : 305 - 336
  • [4] Chand S., 2015, Surv. Oper. Res. Manag. Sci., V20, P35, DOI [DOI 10.1016/J.S0RMS.2015.08.001, 10.1016/j.sorms.2015.08.001, DOI 10.1016/J.SORMS.2015.08.001]
  • [5] Chazal F, 2015, PR MACH LEARN RES, V37, P2143
  • [6] Deb K, 2004, ADV INFO KNOW PROC, P105
  • [7] An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting Approach, Part I: Solving Problems With Box Constraints
    Deb, Kalyanmoy
    Jain, Himanshu
    [J]. IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2014, 18 (04) : 577 - 601
  • [8] Topological persistence and simplification
    Edelsbrunner, H
    Letscher, D
    Zomorodian, A
    [J]. DISCRETE & COMPUTATIONAL GEOMETRY, 2002, 28 (04) : 511 - 533
  • [9] Fasy B., 2015, Tda: Statistical tools for topological data analysis
  • [10] CONFIDENCE SETS FOR PERSISTENCE DIAGRAMS
    Fasy, Brittany Terese
    Lecci, Fabrizio
    Rinaldo, Alessandro
    Wasserman, Larry
    Balakrishnan, Sivaraman
    Singh, Aarti
    [J]. ANNALS OF STATISTICS, 2014, 42 (06) : 2301 - 2339