Bayesian Analysis Made Simple - Woodward, Phil

Résumé :

From simple NLMs to complex GLMMs and beyond, this book describes how to use the GUI for WinBUGS-BugsXLA-an Excel add-in written by the author that allows a range of Bayesian models to be easily specified. With case studies throughout, the text shows how to routinely apply even the more complex aspects of model specification, such as GLMMs, outlier robust models, random effects Emax models, auto-regressive errors, and Bayesian variable selection. It provides brief, up-to-date discussions of current issues in the practical application of Bayesian methods. The author also explains how to obtain free software to get started immediately.

Biographie:

Phil Woodward was born in 1962 in Ipswich, England. After studying Statistics and Mathematics at Brunel University he joined Rolls-Royce in Derby as a statistician in their Nuclear Division. During this time he studied part-time towards a research degree in which he was introduced to the Bayesian paradigm by the late John Naylor and Sir Adrian Smith. Phil then worked for the now defunct Lucas Automotive Company, initially as the Company Statistician but also in various Quality Management roles. Since 1997 Phil Woodward has worked for Pfizer R&D in the UK. He is currently the Global Head of PharmaTherapeutics Statistics, leading the support to the research and development of new medicines from early in the discovery process up to the first studies in patients. He is the creator of the Excel GUI for WinBUGS, BugsXLA, that greatly simplifies the analysis of data using Bayesian methods. Phil is also an active member of the Royal Statistical Society: he was the 2008 Royal Statistical Society's Guy Lecturer for schools, and is a current member of the Editorial Board of its flagship magazine, Significance.

Sommaire:

Brief Introduction to Statistics, Bayesian Methods, and WinBUGS. BugsXLA Overview and Reference Manual. Normal Linear Models (NLMs). Generalized Linear Models. Normal Linear Mixed Models. Generalized Linear Mixed Models. Emax or Four-Parameter Logistic Non-Linear Models. Bayesian Variable Selection. Longitudinal and Repeated Measures Models. Robust Models. Beyond BugsXLA: Extending the WinBUGS Code. Appendix A: Distributions Referenced in BugsXLA. Appendix B: BugsXLA's Automatically Generated Initial Values. Appendix C: Explanation of WinBUGS Code Created by BugsXLA. Appendix D: Explanation of R Scripts Created by BugsXLA. Appendix E: Known Issues.

Critique:
The author in writing this text has succeeded in making Bayesian analysis relatively simple through a graphical user interface (GUI) for WinBUGS-BugsXLA, which resides within Excel. ... I recommend the book to anyone contemplating the use of Bayesian methods for the first time and already familiar with Excel for storing, summarizing and plotting basic statistical data. The text provides an ideal introduction to Bayesian approaches using Excel and ultimately will encourage the reader to migrate to WinBUGS proper. -International Statistical Review, 80, 2012 this book will benefit ... applied statisticians who are familiar with applying generalised linear models and want to consider the impact of bringing Bayesian analyses into their work. ... book will help a competent statistician to run a Bayesian analysis of a generalized linear mixed model almost effortlessly. -John Paul Gosling, Journal of Applied Statistics, 2012