Bayesian Optimization for Quality Diversity Search With Coupled Descriptor Functions

作     者：Kent, Paul Gaier, Adam Mouret, Jean-Baptiste Branke, Juergen 

作者机构：Univ Warwick Coventry CV4 7AL England Autodesk Res AI Lab D-53111 Bonn Germany Univ Lorraine CNRS Inria Loria F-54000 Nancy France Univ Warwick Warwick Business Sch Coventry CV4 7AL England 

出 版 物：《IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION》 (IEEE Trans Evol Comput)

年 卷 期：2025年第29卷第2期

页      面：302-316页

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：Engineering and Physical Sciences Research Council [EP/L015374/1] European Union France 2030 Program through Project PEPR O2R AS3 [ANR-22-EXOD-007] 

主　　题：Optimization Bayes methods Conductivity Linear programming Search problems Evolutionary computation Task analysis Bayesian optimization (BO) quality diversity 

摘      要：Quality diversity (QD) algorithms, such as the multidimensional archive of phenotypic elites (MAP-Elites), are a class of optimization techniques that attempt to find many high-performing points that all behave differently according to a user-defined behavioral metric. In this article we propose the Bayesian optimization of elites (BOP-Elites) algorithm. Designed for problems with expensive fitness functions and coupled behavior descriptors, it is able to return a QD solution-set with excellent performance already after a relatively small number of samples. BOP-Elites models both fitness and behavioral descriptors with Gaussian Process surrogate models and uses Bayesian optimization strategies for choosing points to evaluate in order to solve the quality-diversity problem. In addition, BOP-Elites produces high-quality surrogate models which can be used after convergence to predict solutions with any behavior in a continuous range. An empirical comparison shows that BOP-Elites significantly outperforms other state-of-the-art algorithms without the need for problem-specific parameter tuning.