以下内容来自网站论坛: http://cid-bd8e52b9b639784e.spaces.live.com/blog/cns!BD8E52B9B639784E!147.entry 再议统计生统经典书目 发现本科的统计即使学完了也非常粗浅,更何况没有好好学。下决心好好看一些大师之作。 Probability & Measure: Probability Theory: Theory and Examples, 3rd edition, Richard Durrett 国内有第 2 版影印本 Probability and Measure, Patrick Billingsley Convergence of Probability Measures, Patrick Billingsley A Course in Probability Theory Revised, Kai Lai Chung Mathematical Statitsics: Introduction to Mathematical Statistics, Hogg & Craig (高教社出了第 5 版影印本) Mathematical Statistics, Jun Shao Mathematical Statistics, Peter J. Bickle 作为数理统计学的课本不错。茆诗松、王静龙的《高等数理统计》是国内用得很多的课本。 Inference: Statistical Inference, Casella & Berger 是国外读统计基本必修的的一本书,国内有影印本。 All of Statistics: A Concise Course in Statistical Inference, Larry Wasserman 是一本涵盖面很广 的速成式的 lecture notes 样式的书,偏 nonparametric。Theory of Statistics, Schervish 偏 Bayesian 和 decision theory。此外还有: Testing Statistical Hypotheses, Lehmann & Romano, Theory of Point Estimation, Lehmann & Casella。更多的 advanced topics 举不胜举。 Asymptotics & large sample theory: A Course in Large Sample Theory, Ferguson 是很好的教本, Asymptotic Statistics, A. W. van der Vaart Elements of Large Sample Theory, Lehmann Approximation Theorems of Mathematical Statistics, Serfling Linear Models & Regression: Applied Linear Statistical Models, Kutner et al 或者 Introduction to Linear Regression Analysis, 3ed. Montgomery, Peck, Vining 可以作为入门的 Regression 的教本。 C. R. Rao 的 Linear Statistical Inference and Its Application 很值得看一看。Linear Regression Analysis, Seber & Lee 写得也不错。国内写得很不错的教本是王松桂写的《线性模型引论》, 科学出版社,但是稍有一些错误。 Generalized Linear Models 数 McCullagh & Nelder 最经典,入门可以用 An Introduction to Generalized Linear Models, 3ed, Dobson & Barnett。 Generalized Linear Models: A Bayesian Perspective, Dey, Ghosh, Mallick Categorical Data Analysis, Agresti 是 Categorical 的经典。SAS 和 R 做 Categorical 的手册都有 出版,对于应用统计的 research 来说 Categorical 是很基本的。 Generalized, Linear, and Mixed Models, McCulloch & Seale Linear Mixed Models for Longitudinal Data, Verbeke & Molenberghs Semiparametric Regression, Ruppert, Wand, Carroll 里面把 nonpar & semipar 和 mixed model 统 一起来 SAS for Mixed Models, 2ed 和 Mixed Effects Models in S and S-Plus, Pinheiro & Bates 实现 An Introduction to Multivariate Statistical Analysis, T.W. Anderson Aspects of Multivariate Statistical Theory, Muirhead Applied Multivariate Statistical Analysis, 6ed, Johnson and Wichern 国内有第 6 版影印本 Bayesian Data Aanlysis, Gelman, Carlin, Stern, Rubin Bayesian Methods for Data Analysis, 3rd edition, Carlin & Louis Statistical Decision Theory and Bayesian Analysis, James O. Berger Theory of Statistics, Schervish The Bayesian Choice, Christian P. Robert Bayesian Theory, Bernardo & Smith Generalized Linear Models: A Bayesian Perspective, Dey, Ghosh, Mallick MCMC 比较好的书有 Markov Chain Monte Carlo in Practice, Gilks & Richardson & Spiegelhalter Monte Carlo Strategies in Scientific Computing, Jun S. Liu Monte Carlo Statistical Methods, Robert & Casella Nonparametric & Semiparametric: Semiparametric Regression, Ruppert, Wand, Carroll Applied Nonparametric Regressions, Wolfgang Härdle Nonparametric and Semiparametric Models, Wolfgang Härdle et al Efficient and Adaptive Estimation for Semiparametric Models, Bickel et al All of Nonparametric Statistics, Larry Wasserman Nonparametrics: Statistical Methods Based on Ranks, Erich L. Lehmann Generalized Additive Models, Hastie & Tibshirani 此外,推荐 Empirical Likelihood, Owen Analysis of Longitudinal Data, Diggle, Heagerty, Liang, Zeger Applied Longitudinal Analysis, Fitzmaurice, Laird, Ware Linear Mixed Models for Longitudinal Data, Verbeke & Molenberghs Missing Data & Causal Inference 比较好的书有 Statistical Analysis with Missing Data 2ed, Little & Rubin Missing Data in Clinical Studies, Molenberghs & Kenward Semiparametric Theory and Missing Data, Tsiatis Applied Bayesian Modeling and Causal Inference from Incomplete-Data Perspectives 这本书是 Rubin 这派几代人对 causal inference 的一个集成 Robins 这派的东西就看 paper 吧,Robins & Morgan 09 年也会出一本书 Unified Methods for Censored Longitudinal Data and Causality, Van der Laan & Robins Missing Data in Longitudinal Studies, Daniels & Hogan Bradley Efron 写的 2 本书和 Peter Hall 的 1 本是 Bootstrap 的经典, Bootstrap Methods and Their Application, Davison & Hinkley Survival Analysis: Counting Processes and Survival Analysis, Fleming & Harrington 非常难读但是确实非常经典 Survival Analysis: Techniques for Censored and Truncated Data. Klein & Moeschberger Modeling Survival Data: Extending the Cox Model, Therneau & Grambsh 有大量的 technique The Statistical Analysis of Failure Time Data, Kalbfleisch & Prentice 第一版里处处有一些很好 的 idea,第二版又全面总结了近年的 research development Modelling Survival Data in Medical Research, 2nd edition, David Collett SAS 实现用 Survival Analysis Using the SAS: A Practical Guide, Paul D. Allison subtopics: Analysis of Multivariate Survival Data, Hougaard Bayesian Survival Analysis, Ibrahim, Chen, Sinha Functional Data Analysis, Ramsay & Silverman Elements of Statistical Learning, Hastie, Tibshirani, Friedman Xiao-Hua Zhou 的 Statistical Methods in Diagnostic Medicine 和 Margaret Sullivan Pepe 的 The Statistical Evaluation of Medical Tests for Classification and Prediction 是 Diagnostic Test 的经 典。也有人用 Bayesian 方法做 Clinical Trials 的,可以在 Bayesian Approaches to Clinical Trials and Health-Care Evaluation 里面看到,作者是 Spiegelhalter, Abrams, Myles. 一本极好的书叫《女士品茶:20 世纪统计怎样变革了科学》(The Lady Tasting Tee: How Statistics Revolutionized Science in the Twentieth Century),希望学习统计的都去看看。里面从 Francis Galton, Karl Pearson, R. A. Fisher,讲到 Jerzy Neyman, Emil J. Gumbel, Abraham Wald,Wassily Hoeffding, John Tukey, Florence David Nightingale, Frank Wilcoxon, L.J. Savage, George W. Snedecor, William Cochran, Gertrude Cox, Samuel Wilks, Henry Carver, I.J. Good, Bradley Efron, George Box 的经历. 内容多为一些统计思想的原型,还有许多奇闻轶事。