and it applications. The book is structured in three parts and eighteen chapters. Chapter 1 starts with the main concepts of the Bayes’ Rule and definition of the terms such as likelihood, or probability of evidence given alternative, and base rate, or prior probability of the alternative under consideration. Part I is called likelihood and deals with the evidence in different alternatives. Chapter 2 describes examples of evidence on a car’s color provided to police, and evidence of DNA and of other kinds given for court with the car example, and presents data in tables called Bayes grid and yielding the final estimates for belief, or posterior probability of the alternative conditional to the given evidence. It describes the work of Thomas Bayes with his famous formula of the belief proportional to the likelihood and base rates, and the later works of Laplace on this topic. Chapter 4 suggests other examples of prediction in football and change in GDP in several countries. Chapter 5 provides description for Monty Hall Paradox on which is probability to win a car hidden behind one of three doors, and discusses what Marilyn vos Savant advised for the optimal strategy. Chapter 6 considers opinions in Scotland referendum on splitting from the United Kingdom, evaluated in Bayes grids, with margin of error. Chapter 7 rounds off the likelihood descriptions with examples including the case of letters with anthrax, the assassination of President J.F. Kennedy with consequent death of Oswald and Ruby, and other examples. Part II is called base rate and it considers another part of information needed for Bayesian estimations. Chapter 8 introduces sensitivity and specificity characteristics for 2 × 2 contingency tables on medical examples, and calculates the belief as the product of the base rate and likelihood (see also Lipovetsky and Conklin, 2019). Chapter 9 introduces the characteristics of information and entropy in relation to probability, due to the works by Shannon and von Neumann, Chapter 10 concerns the base rates in the ratio of the alternative and evidence entering into the belief evaluation, considers works by Tversky and Kahneman on the representativeness and availability heuristics, and base rates fallacy, discussed on multiple examples. Chapter 11 discusses a concept of statistical independence by GDP and banks forecasts, and other examples on naïve Bayes classifiers. Chapter 12 reviews and upgrades consideration of examples in the previous chapters. Part III is devoted to application of Bayes approach to more complicated problems. Chapter 13 presents a problem of authorship of the so-called Federalist Papers, assigned mostly to Alexander Hamilton or to James Madison and considered via specifics of words frequencies (see also Mosteller and Wallace, n.d.). Chapter 14 describes estimation in forestry for height of trees by their diameter, and the related problems. Chapter 15 focuses on radiocarbon dating in archeology, particularly, for the image on the Turin Shroud, and for the ancient civilizations in the region around Central America. Chapter 16 discusses various aspects of the Bayesian evidence in the law trials, with multiple examples. Chapter 17 considers the selection tasks often used in psychology experiments, related to reasoning in formal logic and inferencing of the results, where Bayesian rationality helps in correct thinking. Chapter 18 indicates various ways of extension simple Bayes tool to more complicated estimations, including Markov Chain Monte Carlo implemented via winBugs software and Bayesian network analysis, and concludes that “Bayes’ rule is the natural way to think about how to evaluate evidence and use it to revise belief ” (p. 223). Examples in all chapters are given in funny and interesting narratives, historical anecdotes, and stories from life of great scientists. All material is very profound in meaningful description of the problems, there are multiple illustrations and graphs, and each chapter suggests additional references. Numerous cases explain in detail various applications of Bayesian tool which can help to instructors and students, and self-educating readers as well. Tables of Bayes grid in the book show that Bayesian evaluations could be easily and conveniently implemented in Excel spreadsheet. The formal presentation appears only in the appendix of the last two pages: the likelihood, or probability p(E|A) of evidence E given alternative A, the base rate, or prior probability of the alternative p(A), the base rate of evidence p(E), and the belief, or posterior probability p(A|E) (errata: definition for belief in p. 226 should be read as probability of alternative given evidence), and eventually the Bayes formula is presented as p(A|E) = p(E|A)p(A)/p(E) in the last rows of the book.

中文翻译：

---

样本调查和客户满意度数据的参数和非参数统计

和它的应用。本书由三部分十八章构成。第 1 章从贝叶斯规则的主要概念和术语的定义开始，例如可能性，或给定备选方案的证据概率，以及正在考虑的备选方案的基准率或先验概率。第一部分称为可能性，处理不同备选方案中的证据。第 2 章描述了提供给警察的汽车颜色证据的例子，以及提供给法庭的 DNA 证据和其他种类的证据，并以汽车为例，并在称为贝叶斯网格的表格中提供数据，并产生对信念或后验概率的最终估计以给定证据为条件的替代方案。它用他著名的信念与可能性和基本率成正比的公式描述了托马斯贝叶斯的工作，以及拉普拉斯后期关于这个主题的著作。第 4 章提出了几个国家的足球预测和 GDP 变化的其他例子。第 5 章描述了蒙蒂霍尔悖论（Monty Hall Paradox），即赢得隐藏在三扇门之一后面的汽车的概率，并讨论了玛丽莲·沃斯·萨凡特 (Marilyn vos Savant) 对最佳策略的建议。第 6 章考虑了苏格兰公投中关于脱离英国的意见，在贝叶斯网格中进行评估，存在误差。第 7 章用例子完善了可能性描述，包括带有炭疽的信件、肯尼迪总统被暗杀导致 Oswald 和 Ruby 死亡，以及其他例子。第二部分称为基本率，它考虑了贝叶斯估计所需的另一部分信息。第 8 章介绍了医学示例的 2 × 2 列联表的敏感性和特异性特征，并将置信度计算为基本率和可能性的乘积（另请参见 Lipovetsky 和 ​​Conklin，2019）。第 9 章介绍了与概率相关的信息和熵的特征，由于 Shannon 和 von Neumann 的著作，第 10 章涉及进入信念评估的替代和证据比率的基本比率，考虑 Tversky 和 ​​Kahneman 的著作关于代表性和可用性启发式以及基本费率谬误，在多个示例中进行了讨论。第 11 章讨论了 GDP 和银行预测的统计独立性概念，以及其他关于朴素贝叶斯分类器的例子。第 12 章回顾并升级了对前几章示例的考虑。第三部分致力于将贝叶斯方法应用于更复杂的问题。第 13 章提出了所谓的联邦党人论文的作者问题，主要分配给亚历山大·汉密尔顿或詹姆斯·麦迪逊，并通过词频的细节进行考虑（另见莫斯特勒和华莱士，nd）。第 14 章描述了林业中通过直径估算树高的相关问题。第 15 章侧重于考古学中的放射性碳测年，特别是都灵裹尸布的图像以及中美洲周边地区的古代文明。第 16 章讨论了贝叶斯证据在法律审判中的各个方面，并提供了多个例子。第 17 章考虑了心理学实验中经常使用的选择任务，与形式逻辑推理和结果推断有关，贝叶斯理性有助于正确思考。第 18 章指出了将简单贝叶斯工具扩展到更复杂估计的各种方法，包括通过 winBugs 软件和贝叶斯网络分析实现的马尔可夫链蒙特卡罗，并得出结论“贝叶斯规则是思考如何评估和使用证据的自然方式”修正信念”（第 223 页）。所有章节中的例子都以有趣和有趣的叙述、历史轶事和伟大科学家的生活故事给出。所有材料都对问题的有意义的描述非常深刻，有多个插图和图表，每章都提供了额外的参考资料。大量案例详细解释了贝叶斯工具的各种应用，对教师和学生以及自学读者都有帮助。书中的贝叶斯网格表表明，贝叶斯评估可以在 Excel 电子表格中轻松方便地实现。正式表述仅出现在最后两页的附录中：给定备选方案 A 的证据 E 的可能性或概率 p(E|A)，基本率，或备选方案 p(A) 的先验概率，基本率证据 p(E) 和置信度，或后验概率 p(A|E)（勘误：p.226 中置信度的定义应被解读为备选给定证据的概率），最终贝叶斯公式表示为 p (A|E) = p(E|A)p(A)/p(E) 在书的最后几行。