Customer Reviews

Monte Carlo Statistical Methods (Springer Texts in Statistics)

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

 

 

This text may or may not be the best book on MC for a particular application; to be honest, it's the only book on MC I own.

However, I did peruse a number of texts before I bought this one, and I am very pleased with my decision. To me, this book does something that seems necessary but is relatively uncommon: it gives a detailed, modern, comprehensive introduction to MC methods per se. There are other texts that might have one of those characteristics, but they seem to either not have all of them: they either are not modern, not comprehensive, not introductory, or are not concerned with Monte Carlo per se.

Many other excellent texts, for example, are largely oriented toward Bayesian implementations, or general integration, but not both.

I would highly recommend this book as an excellent introduction to MC methods as a general computational tool.

Monte Carlo methods are old. They can be traced back to Buffon's needle problem in the 17th century. However meaningful application had to wait for the invention of digital computers in the 20th century. Much of the development took place in the 1940s and 50s for military and nuclear engineering application. The Hastings - Metropolis algorithm of the 1950s has had a rebirth in the 1990s with the application of Markov Chain Monte Carlo methods to imaging problems and many Bayesian problems.
The authors of this book are Bayesians and present Bayesian methods in the very first chapter. The book is intended to be a course text on Monte Carlo methods. I judge the level to be intermediate to advanced (first or second year graduate level). The first chapter introduces statistical and numerical problems that Monte Carlo methods can solve. It includes a discussion of bootstrap methods in the notes at the end of the chapter. Chapters 2 and 3 introduce standard topics including methods for generating pseudo-random numbers and various variance reduction techniques. Chapter 4 is an introduction to Markov Chains. Markov Chains are commonly a topic in introductory courses on stochastic processes. The authors presuppose that the reader has no knowledge of Markov Chains. So they develop the essential aspects of the theory needed in the application of Markov Chain Monte Carlo methods (MCMC). Chapter 5 then deals with optimization problems discussing simulated annealing, stochastic approximation and the EM algorithm. Chapters 6 - 8 deal with topic in MCMC methods. The final chapter deals with applications to missing data models. The topics are very current and important to statisticians. The theory is covered very well. Many interesting examples are provided throughout the book. A number of these are presented in the problems section at the end of the chapters. It also contains a very extensive bibliography.

I own both versions of this book. The authors have made significant amount of changes and enrichments in the second edition. Many recent developments in this field, such as perfect sampling, trans-dimensional MCMC and sequential Monte Carlo are covered in certain details. The level of this book is intermediate to advanced, and I used this book for the 3rd year Ph.D. students. My only disappointment is the examples are not up to my expectation. However, the problems at the back of each chapter include some interesting applications.
I highly recommend this book to anyone who wants to understand and apply MCMC and other Monte Carlo methods.

There is no doubts this text is a comprehensive study of Monte Carlo methods with an impressive number of examples. However, I must say it is hard to read for someone who is beginning to work with Monte Carlo methods. I highly recommend the book by Sobol (A primer for the Monte Carlo Method) which I think it remains to be the best introduction to the subject. After reading and enjoying this primer you will be ready to take full advantage of Robert and Casella's book.

If you want to understand the theory of MCMC, buy it. (If you also want to understand the theory of stochastic processes, buy Karlin and Taylor (both books - used - they are still the best - but be ready to work) and Parzen (also used)). Then buy "Introducing Monte Carlo Methods With R" (Robert and Casella) and "Bayesian Computation With R" (Albert) to understand how to do MCMC and what it means. Robert is (probably) the best statistician in Europe and one of the best in the world. He also writes extremely well. So does Albert.

This review is about the material quality of the printing in the copy I received. This is not about the content.

I have access to a real copy of this edition in the local library. It is the usual high quality hardcover: it has a matte cover with texture, beautifully bound; the paper inside is high-quality, very soft and slightly off-white; and the printing of the text is very sharp. The version I received from Amazon claimed to be exactly the same, but was very different:
- The hardcover was shiny, did not have texture, and had a natural tendency to bend strongly outwards, it even cannot stay opened if I leave it alone, it will close.
- The paper inside is whiter, horribly white, like standard printing A4 paper;
- The text printing looks like a cheap photocopy of the original. It don't even match a home laser printer. Some formulas are difficult to read. Moreover, some pages are not even centered.

It looks and feels like a cheap knock-off photocopy done in a garage. When I pay a lot of money for a hardcover edition I want the real thing, not a cheap knock-off. Authors should avoid their work being degraded with this cheap printing.

I was trying to read this book in details on importance sampling. It wasted me a few hours looking at the detailed mathematically formula in the corresponding section in the book without getting a clear high level picture. The convoluted examples given the section is more than necessary. Eventually, I found this series of video lecture from mathematicalmonk on youtube.

If you compare the book with this series of video, I believe you will agree this book diverse a 2 star. Technically this book may be sophisticated. But just by sampling the important sampling section and checking a few other sections in the book, I think that I can conclude fairly safely that if there is anything that a reader don't understand in the book, it is the author's fault but not the readers.
[...]
(ML 17.5) Importance sampling - introduction
and the 2 follow up videos

A good book, with a really interesting mathematical treatement to different simulation techniques, but a little bit complicated in some aspects.

It is a fantastic book for Monte Carlo Methods