Customer Reviews

Bootstrap Methods and their Application (Cambridge Series in Statistical and Probabilistic Mathematics)

5 star:		(5)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

This book is loaded with good text book examples and covers a wide variety of bootstrap applications. It is great as a reference book on the bootstrap or as a course text at a graduate level. Chernick (1999) is a little more up-to-date and covers the classifcation error rate estimation problem that is not addressed in this text. Chernick (1999) also has many more references. Efron and Tibshirani (1993) is another fine text that is a little more intuition based with less mathematics. Fieller's problem with ratio estimation and some other gems are well covered in Efron and Tibshirani but not here. Davison and Hinkley do the best job on time series of any of the bootstrap books with details about moving block bootstrap and some interesting applications. The second edition of my text Chernick (2007) just came out and is far more up to date with an improved treatment of time series. But Davison and Hinkley is still by far the best book for a course because of the many exercises and practicals.

This is a well-written book, and I had the basic bootstrap notion figured-out and implemented within a few days, though it will be some time before I develop any serious depth of mastery of the material. For those unfamiliar with the bootstrap method, by a system of resampling from an existing sample of data it provides a means of establishing the standard error for pretty-near any statistical measure (like the standard error of the mean in traditional statistics), as well as the determination of general confidence intervals for those measures, even when the distribution of the data is non-gaussian or unknown. I do find the author's symbolic notation a bit confusing - perhaps TOO compact, and many of the symbolic expressions would really benefit from an associated clearly written paraphrasing (difficult for me to remember all the conventions after putting the book down for a few days). Still, to go from being barely aware of the technique to applying it to the data analysis in a current research project over a period of several weeks suggests that this text does a heck of a good job at conveying the intended introduction.

People might have many different reasons to use bootstrap techniques and they might have various backgrounds therefore this is a good book for readers that want to learn the bootstrap. If you do not have strong background in statistics then the practicals section will help you to programm and analyze the data. If you have background this book does show you the theory behind.

This book has an excellent ballance between practice and theory. It presents the bootstrap as the powerful tool it is through the ellucidation of practical issues. I strongly recommend this book for everyone interested in improving statistical practice.

Relevant and clear in explanations throughout.