Stochastic Orders
Stochastic Orders
复制标题
DOI：
10.1198/tech.2008.s910
复制
发表时间：
2008-08
期刊：
影响因子：
2.5
通讯作者：
中科院分区：
文献类型：
--
作者：
已经看到了SEM方法的快速发展，例如，处理非线性项，多个层次，混合物，等等。这本书提出了国家的最先进的SEM主要从贝叶斯的观点。作者的目的是介绍一种贝叶斯方法，用于在SEM中开发高效和严格的统计方法，并将其应用于实际问题。这本书的目的是作为研究人员的参考和几个学科的研究生的教科书。这本书有几个吸引人的特点。介绍是清晰和明确的，在扫描电镜中的各种主题的讨论是彻底和最新的，和一个专家交织的理论和例子在每个主题使这本书成为一个愉快的阅读和学习。主题的多样性和它们所涵盖和引用的深度使其成为与SEM相关的任何主题研究的非常有用的资源。正如作者所言，“在实践中，经常会遇到有序的分类数据与缺失数据、层次数据和/或异质数据，因此，发展合理的统计方法来科普这种实际情况是有用的。很明显，在整本书中，“健全的统计方法”一直是作者心目中最重要的。基于正态性假设的标准SEM和实现这种SEM的计算机软件具有有限的能力，并且不足以分析复杂的数据集。因此，本书中介绍的贝叶斯发展应该对任何使用SEM的严肃研究人员都有吸引力。在介绍性章节之后，第2章介绍了一些基本的SEM，并对主题进行了温和的介绍。第3章是一个非常规的章节，提供了关于SEM的一些非贝叶斯方法的严格技术细节。这是一个值得欢迎的发展，因为在面向应用的文本中，必要的严谨性往往被回避。第4章和第5章向读者介绍了SEM的贝叶斯观点。其余章节讨论了SEM的不同框架及其分析。这本书是一个受欢迎的除了任何图书馆，应该是一个宝贵的资源，为研究和教学。
have seen rapid development of SEM methodology dealing with, for example, nonlinear terms, multiple levels, mixtures, and so on. This book presents the state-of-the-art SEMs from mainly a Bayesian viewpoint. The author aims to introduce a Bayesian approach for developing efficient and rigorous statistical methodologies in SEMs and applying them to practical problems. The book is intended to serve as a reference for researchers and a textbook for graduate students in several disciplines. The book has several attractive features. The presentation is lucid and clear, the discussion of various topics in SEM is thorough and up to date, and an expert interweaving of theory and examples in each topic make the book a pleasure to read and learn from. The variety of topics and the depth to which they are covered and referenced makes this an extremely useful resource for research in any topic relating to SEMs. As the author remarks, “it is very common in practice to encounter ordered categorical data with missing data, hierarchical data and/or heterogeneous data, and hence developments of sound statistical methods to cope with such practical situations are useful.” It is clear throughout the book that “sound statistical methods” have been topmost in the author’s mind. The standard SEMs based on normality assumption and computer software implementing such SEMs have limited capabilities and are inadequate for analyzing complex data sets. Thus the Bayesian development presented in this book should be attractive to any serious researcher using SEMs. After the introductory chapter, Chapter 2 presents some basic SEMs and provides a gentle introduction to the subject matter. Chapter 3 is an off-beat chapter that provides rigorous technical detail about some non-Bayesian methods of SEM. This is a welcome development, because necessary rigor often is sidestepped in application-oriented texts. Chapters 4 and 5 introduce the reader to the Bayesian viewpoint of SEMs. The remaining chapters discuss different frameworks for SEM and their analysis. This book is a welcome addition to any library and should be a valuable resource for research and teaching.