Optimal design of experiments subject to correlated errors

被引：24

作者：

Pázman, A

Müller, WG

机构：

[1] Comenius Univ, Bratislava 84215, Slovakia

[2] Vienna Univ Econ & Business Adm, A-1090 Vienna, Austria

来源：

STATISTICS & PROBABILITY LETTERS | 2001年 / 52卷 / 01期

关键词：

design measure; approximate information matrix; random field; exact designs;

D O I：

10.1016/S0167-7152(00)00201-7

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 paper we consider optimal design of experiments in the case of correlated observations, when no replications are possible. This situation is typical when observing a random process or random field with known covariance structure. We present a theorem which demonstrates that the computation of optimum exact designs corresponds to solving minimization problems in terms of design measures. (C) 2001 Elsevier Science B.V. All rights reserved.

引用

页码：29 / 34

页数：6

共 7 条

[1]

[Anonymous], OPTIMAL DESIGN EXPT

[2]

BRIMKULOV U, 1980, ZAVODSK LAB, V36, P435

[3]

Kiefer J., 1959, STAT MODELS LINEAR S, P272

[4] An algorithm for the computation of optimum designs under a given covariance structure [J].

Müller, WG ;

Pázman, A .

COMPUTATIONAL STATISTICS, 1999, 14 (02) :197-211

[5]

MULLER WG, 1998, CONTRIBUTIONS STAT

[6]

PAZMAN A, 1998, MODEL ORIENTED DATA, V5, P239

[7] DESIGNS FOR REGRESSION PROBLEMS WITH CORRELATED ERRORS [J].

SACKS, J ;

YLVISAKE.D .

ANNALS OF MATHEMATICAL STATISTICS, 1966, 37 (01) :66-&

← 1 →