If you consider yourself Econometrician\Statistician or one of those numerous buzz word synonyms that are floating around these days, Computer Age Statistical Inference: Algorithms, Evidence and Data Science by Bradley Efron and Trevor Hastie is a book you can’t miss, and now nor should you. You can download the book for free.
My first inclination is to deliver an unequivocal recommendation. But in truth, my praises would probably fall short of what was already written.
So what can I give you? I can say that there are currently 6 amazon reviews, with a 4.5 average. One of the reviewers writes that there is some overlap with previous work. I agree. But it doesn’t matter. It reads so well, call it a refresher. Let’s face it, it is not as if you always have it so clear in your head such that you can afford to skip sections because you read something similar before.
I can also tell you why I think it is a special book.
In my opinion most scientists who are as devoted, passionate and knowledgeable around a math-heavy field are not very good at communicating with people. I think many fields could benefit from having their top researchers communicate better with their interested community. The thing is, the more you dive into complicated math-heavy topics (measure theory comes to mind), the more you dive into those topics. Also, because they are math-heavy they are not easily linked to real-life, unlike behavioral sciences for example. In addition, the more years go by, the easier it becomes for the researcher to forget how hard it was for him\her picking the knowledge. So, what we need for the creation of such a book are researchers who have devoted their life to study those complicated topics. Researchers who still recognize the importance of communicating their decades of study with people outside the academic world. Researchers who still have the sensitivity to identify what is well understood, and what needs emphasis. In the field of Statistics (or one of those numerous buzz word synonyms that are floating around these days), there are only a handful of those researchers around. Bradley Efron and Trevor Hastie are two of those. This is what makes this book so special.
I somehow don’t think the table of contents reveals the spirit of the book. The book takes you along the path of the development of Statistics over the different generations. Discussions of classical (Fisher and Neyman–Pearson) inference, Bayesian statistics, bootstrapping and machine learning algorithms, and finally large scale inference (FDR). So the historical perspective is delightful to read and is well placed in context. As the book progresses in its chronology, you will get some top-notch descriptions of advanced methods. The kind of descriptions that can only be written by this kind of giants. The word description is intentional here, it is not as hands-on as other, more-specialized books- so align your expectations in that regards. Although I dare write that if you have read more specialized books already, those might now become clearer in light of those superb descriptions.
A final word of caution. If you encounter a formula where you think something should be transposed instead of inverted- doubt yourself sure, but also check the errata.
In sum, a must-read, for you lovers of statistics.
-_-…
Another book completely full of abstract mathematical notation, without a numerical example or a line of R code in sight.
Trevor Hastie had it so right with intro to statistical learning with R. And it is my opinion that every book he writes have some examples so that non-PhDs can more easily digest the material.
I’m a big fan of the book Introduction to Statistical Learning in R, since it’s well written and also has R code to accompany it. But I don’t see any R code in CASI. The CASI TOC looks great, so I will definitely look into it. Thank you for pointing CASI out in you great blog.