Composite grid designs for adaptive computer experiments with fast inference
Composite grid designs for adaptive computer experiments with fast inference
复制标题
DOI：
10.1093/biomet/asaa084
复制
发表时间：
2021-08
期刊：
影响因子：
2.7
通讯作者：
M. Plumlee;Collin B. Erickson;Bruce E. Ankenman;E. Lawrence
中科院分区：
文献类型：
--
作者：
M. Plumlee;Collin B. Erickson;Bruce E. Ankenman;E. Lawrence
实验常被用于生成确定性计算机代码的模拟器。本文介绍了复合网格实验设计以及一种用于构建精确模拟设计的序贯方法。开发了一些计算方法，即使在样本量很大的情况下，也能实现快速且精确的高斯过程推断。我们证明了所提出的方法能够生成模拟器，其精度比当前的近似方法高出几个数量级，而计算成本相当。
Experiments are often used to produce emulators of deterministic computer code. This article introduces composite grid experimental designs and a sequential method for building the designs for accurate emulation. Computational methods are developed that enable fast and exact Gaussian process inference even with large sample sizes. We demonstrate that the proposed approach can produce emulators that are orders of magnitude more accurate than current approximations at a comparable computational cost.