Order-based error for managing ensembles of surrogates in mesh adaptive direct search

被引：24

作者：

Audet, Charles ^{[1
,2
]}

Kokkolaras, Michael ^{[3
,4
]}

Le Digabel, Sebastien ^{[1
,2
]}

Talgorn, Bastien ^{[3
,4
]}

机构：

[1] Ecole Polytech Montreal, GERAD, CP 6079,Succ Ctr Ville, Montreal, PQ H3C 3A7, Canada

[2] Ecole Polytech Montreal, Dept Math & Genie Ind, CP 6079,Succ Ctr Ville, Montreal, PQ H3C 3A7, Canada

[3] McGill Univ, GERAD, 817 Sherbrooke St West, Montreal, PQ H3A 0C3, Canada

[4] McGill Univ, Dept Mech Engn, 817 Sherbrooke St West, Montreal, PQ H3A 0C3, Canada

来源：

JOURNAL OF GLOBAL OPTIMIZATION | 2018年 / 70卷 / 03期

基金：

加拿大自然科学与工程研究理事会;

关键词：

Derivate-free optimization; Ensemble of surrogates; MADS; Order error; EFFICIENT GLOBAL OPTIMIZATION; RADIAL BASIS FUNCTIONS; MODELS; ALGORITHM; FRAMEWORK; FLOW;

D O I：

10.1007/s10898-017-0574-1

中图分类号：

C93 [管理学]; O22 [运筹学];

学科分类号：

070105 ; 12 ; 1201 ; 1202 ; 120202 ;

摘要：

We investigate surrogate-assisted strategies for global derivative-free optimization using the mesh adaptive direct search (MADS) blackbox optimization algorithm. In particular, we build an ensemble of surrogate models to be used within the search step of MADS to perform global exploration, and examine different methods for selecting the best model for a given problem at hand. To do so, we introduce an order-based error tailored to surrogate-based search. We report computational experiments for ten analytical benchmark problems and three engineering design applications. Results demonstrate that different metrics may result in different model choices and that the use of order-based metrics improves performance.

引用

页码：645 / 675

页数：31

共 75 条

[1] ORTHOMADS: A DETERMINISTIC MADS INSTANCE WITH ORTHOGONAL DIRECTIONS [J].