A stochastic characterization of Loewner optimality design criterion in linear models

被引：6
|
作者：
Liski, EP
Zaigraev, A
机构：
[1] Univ Tampere, Dept Math Stat & Philosphy, FIN-33101 Tampere, Finland
[2] Nicholas Copernicus Univ, Fac Math & Comp Sci, PL-87100 Torun, Poland
来源：
METRIKA | 2001年 / 53卷 / 03期
关键词：
Anderson's theorem; stochastic convex and distance optimality criteria; Kiefer optimality; admissibility; information equivalence; multifactor first degree polynomial models; orthogonal and simplex designs;
D O I：
10.1007/s001840000106
中图分类号：
O21 [概率论与数理统计]; C8 [统计学];
学科分类号：
020208 ; 070103 ; 0714 ;
摘要：
In this paper we present a new stochastic characterization of the Loewner optimality design criterion. The result is obtained by proving a generalization to the well known corollary of Anderson's theorem. Certain connections between the Loewner optimality and the stochastic distance optimality design criterion are showed. We also present applications and generalizations of the main result.
引用
收藏
页码：207 / 222
页数：16
相关论文
共 50 条
12345