Experimental design in a class of models
作者:
Elisabeth Läuter,
期刊:
Mathematische Operationsforschung und Statistik
(Taylor Available online 1974)
卷期:
Volume 5,
issue 4-5
页码: 379-398
ISSN:0047-6277
年代: 1974
DOI:10.1080/02331887408801175
出版商: Akademie-Verlag
数据来源: Taylor
摘要:
Let there be a setFof mathematical modelsLm(x) for the description of the depend once of a real random variabley(x) on, wheremis an element of a subscript setMwhich may be continuous. Let. We suppose that the models are linearly dependent on the parameters. The observation errors ε(x) may be uncorrelated for different measurements, and. The problem is to find an experimental design which enables us to find the best possible estimate of the parameters for all models ofF. Therefore the following criteria are used: For a measureQ,on a σ algebra ofMwe consider the functionalsandwhereis the determinant of the information matrix of the modelLmfor the continuous design ξ. Properties of designs maximizingC(ξ) andS(ξ) are given. For instance a minimax property, and iterative computing procedures forCandS-optimal estimation and discrimination problems can be treated using the given method. A relationship to the results of S.M. STIGLER (1971) and A.C. ATKINSON (1972) is also indicated.
点击下载:
PDF (967KB)
返 回