SEQUENTIAL DESIGN OF EXPERIMENTS FOR NONLINEAR MODELS.
SEQUENTIAL DESIGN OF EXPERIMENTS FOR NONLINEAR MODELS.
复制标题
非线性模型实验的序贯设计。
DOI：
--
复制
发表时间：
1963
期刊：
影响因子：
--
通讯作者：
W. G. Hunter
中科院分区：
文献类型：
--
作者：
G. Box;W. G. Hunter
摘要：对于工程师来说，基本的机构研究是有意义的，主要是因为更深入的理解使人们有可能以一种比机构完全未知的方式更智能和更有用的方式处理工程设计问题。这种机理研究基本上包括两个步骤：(1)为理论模型建立适当的形式，然后(2)准确地确定其参数的值。在本文中，假设步骤(I)已经完成，从而知道理论模型的形式。现在实验者面临的问题是对物理参数(例如，化学动力学例子中的速率常数)的评估。本文的目的是考虑数据的生成问题，即实验的统计设计问题。
Abstract : To engineers, basic mechanism studies are of interest principally because a deeper understanding makes it possible to cope with engineering design problems in a more intelligent and useful manner than would be possible if the mechanism were entirely unknown. Such mechanism studies consist essentially of two steps: (i) establishing an adequate form for the theoretical model and then (ii) determining precisely the values of its parameters. In this paper it is supposed that step (i) has been accomplished and the form of the theoretical model is therefore known. The problem which confronts the experimenter now is the evaluation of the physical parameters (e.g., rate constants in chemical kinetics examples). The purpose of this paper is to consider the problem of generation of data, i.e., the statistical design of experiments.