内容简介
this book consists of solutions to 400 exercises, over 95% of which arein my book Mathematical Statistics. Many of them are standard exercisesthat also appear in other textbooks listed in the references. It is onlya partial solution manual to Mathematical Statistics (which contains over900 exercises). However, the types of exercise in Mathematical Statistics notselected in the current book are (1) exercises that are routine (each exerciseselected in this book has a certain degree of difficulty), (2) exercises similarto one or several exercises selected in the current book, and (3) exercises foradvanced materials that are often not included in a mathematical statisticscourse for first-year Ph.D. students in statistics (e.g., Edgeworth expan-sions and second-order accuracy of confidence sets, empirical likelihoods,statistical functionals, generalized linear models, nonparametric tests, andtheory for the bootstrap and jackknife, etc.). On the other hand, this isa stand-alone book, since exercises and solutions are comprehensibleindependently of their source for likely readers. To help readers notusing this book together with Mathematical Statistics, lists of notation,terminology, and some probability distributions are given in the front ofthe book.
内页插图
目录
Preface
Notation
Terminology
Some Distributions
Chapter 1. Probability Theory
Chapter 2. Fundamentals of Statistics
Chapter 3. Unbiased Estimation
Chapter 4. Estimation in Parametric Models
Chapter 5. Estimation in Nonparametric Models
Chapter 6. Hypothesis Tests
Chapter 7. Confidence Sets
References
Index
前言/序言
Since the publication of my book Mathematical Statistics (Shao, 2003), Ihave been asked many times for a solution manual to the exercises in mybook. Without doubt, exercises form an important part of a textbookon mathematical statistics, not only in training students for their researchability in mathematical statistics but also in presenting many additionalresults as complementary material to the main text. Written solutionsto these exercises are important for students who initially do not havethe skills in solving these exercises completely and are very helpful forinstructors of a mathematical statistics course (whether or not my bookMathematical Statistics is used as the textbook) in providing answers tostudents as well as finding additional examples to the main text. Moti-vated by this and encouraged by some of my colleagues and Springer-Verlageditor John Kimmel, I have completed this book, Mathematical Statistics:Exercises and Solutions.
This book consists of solutions to 400 exercises, over 95% of which arein my book Mathematical Statistics. Many of them are standard exercisesthat also appear in other textbooks listed in the references. It is onlya partial solution manual to Mathematical Statistics (which contains over900 exercises). However, the types of exercise in Mathematical Statistics notselected in the current book are (1) exercises that are routine (each exerciseselected in this book has a certain degree of difficulty), (2) exercises similarto one or several exercises selected in the current book, and (3) exercises foradvanced materials that are often not included in a mathematical statisticscourse for first-year Ph.D. students in statistics (e.g., Edgeworth expan-sions and second-order accuracy of confidence sets, empirical likelihoods,statistical functionals, generalized linear models, nonparametric tests, andtheory for the bootstrap and jackknife, etc.). On the other hand, this isa stand-alone book, since exercises and solutions are comprehensibleindependently of their source for likely readers. To help readers notusing this book together with Mathematical Statistics, lists of notation,terminology, and some probability distributions are given in the front ofthe book.
数理统计:从基础理论到前沿应用 书籍简介 本书旨在为读者提供一个全面、深入且具有实践指导意义的数理统计学习路径。我们聚焦于统计学核心理论的严谨构建,并辅以大量详实的实例分析,以期帮助读者不仅掌握统计学的“是什么”,更能理解其“为什么”和“怎么用”。全书内容覆盖了从概率论基础到高级推断方法的广阔领域,力求在理论深度与实际应用之间架起一座坚实的桥梁。 第一部分:概率论与随机变量基础 数理统计的根基在于概率论。本部分将系统回顾和深化读者对概率空间、随机变量、概率分布等基本概念的理解。我们从测度论的视角审视概率的定义,确保理论基础的牢固性。 1. 概率空间与随机变量: 详细阐述 $sigma$-代数、可测函数及其在概率空间中的意义。随机变量的定义、分类(离散、连续、混合)及其期望、方差等矩的计算方法得到细致的讨论。特别关注Lebesgue积分与Riemann积分在概率论中的联系与区别。 2. 重要概率分布: 深入剖析常见的单变量和多变量概率分布,如二项分布、泊松分布、正态分布、卡方分布、t分布和F分布。对于多维随机变量,重点分析其联合分布、边际分布、条件期望,以及协方差矩阵的性质。向量值随机变量的特征函数和联合中心极限定理是本节的难点与重点。 3. 大数定律与中心极限定理: 检验样本均值的收敛性和极限分布是统计推断的基础。本书将介绍强大数定律和弱大数定律的不同版本,并提供中心极限定理的多种形式(如Lindeberg-Feller CLT),解析其在统计推断中的普适性。 第二部分:统计推断的基础框架 在概率论的坚实基础上,本部分转向统计推断的核心:如何从有限样本中可靠地提取信息并对总体做出判断。 1. 随机样本与统计量: 明确随机样本的概念,并系统介绍各种常用的统计量,如样本均值、样本方差、矩估计量等。重点讨论统计量的分布,特别是基于正态总体的各种二次型统计量的分布特性。 2. 参数估计理论: 点估计: 详尽介绍矩估计法(MOM)和最大似然估计法(MLE)。对于MLE,我们将深入探讨其渐近性质(一致性、渐近正态性、渐近有效性),并阐述Cramér-Rao下界理论,用以衡量估计量的优劣。 区间估计: 讲解置信区间的构建原理,包括基于枢轴量、delta方法以及Bootstrap方法的构建技巧。重点分析不同置信水平的实际含义和稳健性。 第三部分:假设检验的严谨方法 假设检验是统计决策的核心工具。本部分旨在提供一套系统而严谨的检验流程和评估标准。 1. 检验原理与结构: 界定原假设 ($H_0$) 和备择假设 ($H_1$),详细解释I类错误(显著性水平 $alpha$)和II类错误(功效 $1-eta$)。介绍似然比检验(LRT)作为构建最优检验量的通用框架。 2. 经典检验方法: 涵盖参数假设检验的主要内容,包括: 基于正态性假设的t检验、F检验(方差齐性检验)。 卡方检验(拟合优度检验、独立性检验)。 基于非参数方法的检验,例如Wilcoxon符号秩检验、Mann-Whitney U检验的原理介绍。 3. 检验的性能评估: 不仅关注p值的使用,更强调检验功效的计算和提升。讨论稳健性(Robustness)问题,即当模型假设被轻微违反时,检验的性能如何变化。 第四部分:线性模型的统计推断 线性模型是应用统计学中最强大和最常用的工具之一。本部分侧重于多元数据分析的理论基础。 1. 多元正态分布: 作为多元回归分析的理论前提,对多元正态分布的性质(如边缘分布、条件分布、独立性、二次型的分布)进行详尽的阐述。 2. 一般线性模型(GLM): 深入探讨多元线性回归模型的最小二乘估计(OLS)的性质,包括估计量的无偏性、有效性和分布。重点解析高斯-马尔可夫定理及其在OLS最优性中的地位。 3. 模型诊断与推断: 讲解如何进行系数的显著性检验(t检验)、模型的整体显著性检验(F检验)。深入讨论残差分析、共线性问题(方差膨胀因子VIF)以及异方差性(如White检验)的处理和修正方法。 第五部分:进阶主题与统计计算 为使读者接触现代统计学的最新发展,本部分引入了更复杂的模型和计算方法。 1. 非参数统计的深度探讨: 介绍核密度估计(KDE)的理论基础,包括核函数的选择和带宽(Bandwidth)的优化。讨论经验过程(Empirical Processes)在现代统计推断中的作用。 2. 统计计算与模拟方法: 阐述蒙特卡洛(Monte Carlo)模拟在复杂积分和分布逼近中的应用。详细介绍马尔可夫链蒙特卡洛(MCMC)方法,特别是Metropolis-Hastings算法和Gibbs抽样,它们是贝叶斯统计计算不可或缺的工具。 3. 贝叶斯统计导论: 从频率学派的视角引入贝叶斯框架,讨论先验分布、似然函数和后验分布的构建。介绍如何利用后验分布进行参数估计和区间预测,并对比贝叶斯方法与经典方法的哲学差异和实际优势。 适用对象 本书内容严谨、推导详尽,适合于数学、统计学、物理学、工程学、经济学及生物统计学等专业的高年级本科生、研究生,以及需要深入理解统计学理论基础的科研人员和数据分析师。阅读本书需要具备扎实的微积分和线性代数基础,以及初步的概率论知识。 本书特色 本书的重点在于理论的内在逻辑和推导的完整性。每一个核心概念的引入都伴随着严格的数学证明或清晰的逻辑阐述。我们避免了对复杂计算过程的过度简化,力求呈现一个真实、无捷径的数理统计学习体验。每一章末尾设置了具有挑战性的思考题,用以巩固和拓展读者的理论掌握程度。