Statistics: The Exploration and Analysis of Data
Statistics: The Exploration and Analysis of Data
复制标题
DOI：
10.1198/004017002320256585
复制
发表时间：
2002-08
期刊：
影响因子：
2.5
通讯作者：
P. Minton
中科院分区：
文献类型：
--
作者：
P. Minton
这些思想、迭代算法是为估计所研究的空间过程的一阶特征而提出的最优或接近最优的设计。对于测量可以及时重复的情况，例如在环境监测的情况下，还提出了进行最优设计的方法。第六章讨论了估计空间过程二阶特征的最优设计问题，如变异函数估计。同样，经典理论也适用于处理空间相关性。讨论了几种算法，并对它们的性能进行了比较。提出了一种用于组合对不同目标最优的设计以获得具有良好整体性能的单一设计的方法，尽管这样的设计不一定对任何单个目标是最优的。在每一章中，都使用与重建由17个监测点组成的上奥地利二氧化硫监测网有关的空间数据来说明这一理论。此外，使用第二个真实示例数据集，开发了练习列表。该数据集包括1992-1997年期间下奥地利中部多瑙河流域36个水质监测站的氯化物测量值。所有数据均可从http://statistik.wu-wien.ac.at/stat4/mueller/csd/data.zip.下载在每一章的末尾都有足够的期刊文章和书籍参考文献。有七个附录。附录A.1对本书中使用的两个主要数据集进行了说明。本书正文中提到的各种不平凡的数学事实的证明见附录A.2至A.6。根据作者的说法，书中报告的计算是使用一个名为D2PT的程序完成的，该程序是用GAUSS386V3.2(Aptech，1993)编写的。可以从http://statistik.wu-wien.ac.at/stat4/mueller/csd/d2pt.zip为这本书下载应用程序模块。附录A.7对D2PT中使用的变量进行了说明。本书中使用的符号及其含义的列表紧跟在
these ideas, iterative algorithms are formulated for nding optimum or nearly optimum designs for estimating the rst order characteristics of the spatial process under study. Methods for nding optimum designs are also presented for the case where measurements may be replicated in time, such as in the case of environmental monitoring. Chapter 6 treats the problem of nding optimum designs for estimating second order characteristics of a spatial process such as variogram estimation. Again, classical theory is adapted to handle spatial correlations. Several algorithmic approaches are discussed and their performances compared. A method is proposed for combining designs that are optimum for different objectives to obtain a single design with good overall performance, although such a design may not necessarily be optimal for any of the individual objectives. In each chapter the theory is illustrated using spatial data related to the reconstruction of the Upper-Austrian sulfur dioxide monitoring network consisting of 17 monitoring sites. Additionally, using a second real example dataset, a list of exercises is developed. This dataset consists of Chloride measurements from 36 water quality monitoring stations in the Danube River Basin in central Lower Austria for the period 1992–1997. All of the data may be downloaded from http://statistik.wu-wien.ac.at/stat4/mueller/csd/data.zip. Adequate journal article and book references are given at the end of each chapter. There are seven appendices. Appendix A.1 gives a description of the two main data sets used in the book. Proofs of various nontrivial mathematical facts mentioned in the body of the book are given in Appendixes A.2 through A.6. According to the author, computations reported in the book were done using a program called D2PT which is written using GAUSS386 V3.2 (Aptech, 1993). An application module may be downloaded from http://statistik.wu-wien.ac.at/stat4/mueller/csd/d2pt.zip for the book. A description of the variables used in D2PT is given in Appendix A.7. A list of symbols used in the book and their meanings are given immediately following the Table of