多响应模型判别的试验设计

The Design of Experiments for Discriminating Between Multiresponse Models

Atkinson和Fedorov在1975年给出了对两个单响应线性模型进行判别试验设计的一些结论。基于这些结论,通过Khuri(1985)采用的对多响应模型进行线性组合的方法,针对两组多响应模型判别作出试验设计,并给出了判别准则、等价定理以及构造最优设计的一种序贯方法;最后也对多组多响应模型判别的试验设计作出进一步的讨论。

Atkinson and Fedorov（1975） have given some results for the design of experiments for discriminating between two singleresponse models. Basing on these results, this paper constructs the design of experiments for discriminating between two multiesponse models and deals with the multiesponse models by a way of Khuri（1985）, which do a linear combination to the responses. The design criteria, equivalence theorem and sequential designs are obtained. At last, a discussion of the design for discriminating between several multiresponse models is given.