Chemometrics: A Textbook

Chemometrics: A Textbook
复制标题

DOI:
10.1080/00401706.1990.10484611
复制
发表时间:
1990-02
期刊:
影响因子:
2.5
通讯作者:
G. A. Mack
G. A. Mack
中科院分区:
工程技术3区
文献类型:
--
作者:
G. A. Mack

文献摘要

被引文献

相似文献

总的来说,对于化学家和统计学家来说,我推荐这本书值得一读。对于化学家来说,它可以产生兴趣和理解为什么以及在分析化学的许多应用领域中需要统计方法的地方。对统计学家来说。这本书提供了对化学的许多定量领域的介绍,而统计学家通常不会熟悉这些领域。为一些重要的校准和优化问题提供了有用的见解。然而,我并不推荐这本书作为化学家的“如何”统计参考文献。我觉得没有足够的背景材料让非统计学家真正了解所讨论的统计方法的应用。在很多地方。这种讨论可能会让非统计学家感到困惑。例如,关于最小二乘回归的讨论涵盖了一系列广泛的主题,但过于肤浅,对应用程序来说没有用处。对方差分析(ANOVA)表、均方差和残差分析的介绍需要更多的背景讨论才能对非统计学家有用。排除复杂的细节,简单地发展基本概念,并让感兴趣的读者参考其他来源的其他更详细的讨论,会更有用。我还质疑所提供的许多公式是否都那么有用,因为大多数从业者都会使用现有的计算机程序包进行分析。
Overall, I recommend this book as worthwhile reading for chemists and statisticians. For chemists, it can generate an interest and understanding of why and where statistical methods arc needed in the many application areas of analytical chemistry. For statisticians. the book provides an introduction to many quantitative areas of chemistry with which a statistician would not ordinarily be familiar. Useful insight is given for a number of important calibration and optimization problems. I do not, however, recommend the book as a “how to” statistical rcfercnce document for chemists. I feel that there is insufficient background material to allow nonstatisticians to really understand the application of the statistical methods discussed. In many places. the discussion is likely to be confusing to nonstatisticians. For example, the discussion on least squares regression covers a broad array of topics in much too superficial a manner to be useful for application purposes. The introduction to analysis of variance (ANOVA) tables, mean squares, and analysis of residuals requires more background discussion to be useful to nonstatisticians. It would have been more useful to exclude complicating details and simply develop the fundamental concepts and refer interested readers to other, more detailed discussions in other sources. I also question whether many of the formulas provided are all that useful, because most practitioners will use existing computer packages to conduct analyses.