The mesh adaptive direct search algorithm with treed Gaussian process surrogates
The mesh adaptive direct search algorithm with treed Gaussian process surrogates
复制标题
树状高斯过程代理的网格自适应直接搜索算法
DOI：
--
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
R. Gramacy
中科院分区：
文献类型：
--
作者：
Sébastien Le Digabel;R. Gramacy
这项工作介绍了在网格自适应直接搜索（MADS）框架中使用树状高斯过程（TGP）作为代理模型进行约束黑盒优化。它将代理管理框架（SMF）扩展到一般约束下的非平滑优化。MADS以两种方式使用TGP：一是作为黑盒评估的替代；第二，评估统计标准，如预期的改善和方差的平均减少。在三个问题上验证了该方法的有效性：一个具有多个局部最优解的综合问题；化工模拟器在苯乙烯生产中的一个实际应用；还有一个来自污染物清理和水文学。在所有三种情况下，我们都表明TGP代理比二次模型和没有任何代理的MADS更可取。
This work introduces the use of the treed Gaussian process (TGP) as a surrogate model within the mesh adaptive direct search (MADS) framework for constrained blackbox optimization. It extends the surrogate management framework (SMF) to nonsmooth optimization under general constraints. MADS uses TGP in two ways: one, as a surrogate for blackbox evaluations; and two, to evaluate statistical criteria such as the expected improvement and the average reduction in variance. The eciency of the method is tested on three problems: a synthetic one with many local optima; one real application from a chemical engineering simulator for styrene production; and one from contaminant cleanup and hydrology. In all three cases we show that the TGP surrogate is preferable to a quadratic model and to MADS without any surrogate at all.
影响因子：
2.2
作者：
Gramacy, Robert B.;Lee, Herbert K. H.
通讯作者：
Lee, Herbert K. H.
影响因子：
2.4
作者：
Gramacy, Robert B.;Polson, Nicholas G.
通讯作者：
Polson, Nicholas G.