基于ANOVA高维模型的正交设计的优良性(英文)  

Optimal Properties of Orthogonal Arrays Based on ANOVA High-Dimensional Model Representation

作　　者：陈雪平[1,2] 林金官[1] 王晓迪[3] 

机构地区：[1]东南大学数学系,南京211189 [2]江苏理工学院数学系,常州213001 [3]中央财经大学统计学院,北京100081

出　　处：《应用概率统计》2013年第6期570-580,共11页Chinese Journal of Applied Probability and Statistics

基　　金：supported by Natural Science Foundation of Jiangsu Province(BK2011058);National Natural Science Foundation of China(11171065 and 11301073)

摘　　要：全局敏感性指标在全局敏感性分析中占有重要的地位,Wang等(2012)证明了正交设计在估计参数βM时具有A最优,本文论证了正交设计在估计参数时的一些其他最优性质,包括估计参数βM的E最优和估计参数θM的一致最优性.在模拟论证中,我们提出了用随机化正交表来替代一般的正交表,并得到了较好的性质,如减少了偏差并且提高了精度.Global sensitivity indices play important roles in global sensitivity analysis based on ANOVA high-dimensional representation, Wang et al. （2012） showed that orthogonal arrays are A-optimality designs for the estimation of parameter θM, the definition of which can be seen in Section 2. This paper presented several other optimal properties of orthogonal arrays under ANOVA high- dimensional representation, including E-optimality for the estimation of θM and universal opti- mality for the estimation of βM, where βM is the independent parameters of θM. Simulation study showed that randomized orthogonal arrays have less biased and more precise in estimating the confidence intervals comparing with other methods.

关 键 词：矩阵象 E最优 一致最优 正交设计 

分 类 号：O212.6[理学—概率论与数理统计]