The correct Kriging variance estimated by bootstrapping

被引：56

作者：

den Hertog, D

Kleijnen, JPC

Siem, AYD

机构：

[1] Tilburg Univ, Dept Econometr & Operat Res, Ctr Econ Res, NL-5000 LE Tilburg, Netherlands

[2] Tilburg Univ, Dept Informat Syst & Management, Ctr Econ Res, NL-5000 LE Tilburg, Netherlands

来源：

JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY | 2006年 / 57卷 / 04期

关键词：

Kriging; simulation; bootstrapping; design and analysis of computer experiments ( DACE); global optimization;

D O I：

10.1057/palgrave.jors.2601997

中图分类号：

C93 [管理学];

学科分类号：

12 [管理学]; 1201 [管理科学与工程]; 1202 [工商管理学]; 120202 [企业管理];

摘要：

The classic Kriging variance formula is widely used in geostatistics and in the design and analysis of computer experiments. This paper proves that this formula is wrong. Furthermore, it shows that the formula underestimates the Kriging variance in expectation. The paper develops parametric bootstrapping to estimate the Kriging variance. The new method is tested on several artificial examples and a real-life case study. These results demonstrate that the classic formula underestimates the true Kriging variance.

引用

页码：400 / 409

页数：10

共 22 条

[1]

[Anonymous], [No title captured]

[2]

A rigorous framework for optimization of expensive functions by surrogates [J].