Subscribe to DSC Newsletter

Not a new book, it was published in 1993, and at $202 for 228 pages, it's probably one of the most expensive statistics book per page, especially for such an old book. I have no idea what justifies such a high price. Do you guys see a $202 price on Amazon, or is it just me because I once mentioned that Amazon could try user-customized prices: showing different prices to different users, based on the user's purchase history.

The reason why I mentioned this book is because I bought a used version (at a steeply discounted price) and Chapter 4 (least-absolute-deviations regressions) is closely related to my new correlation coefficient designed for big data. This natural correlation has always been disregarded and not studied by traditional statisticians because it does not lead to elegant and simple mathematical formulas. It's a bit a black sheep construct, because it does not fit with the GLM (General Linear Model) theory.

With the big data revolution and availability of cheap, fast and clustered computer processors, things have changed (I guess not for traditional statisticians still practicing cave-man statistics), and whether an elegant mathematical formula exists or not, is the least of our concerns. Surprisingly, an elegant formula actually does exist in this non-standard framework: two scientists, optimization experts - a distinguished engineer from IBM France, and a chief scientist at John Hopkins - have actually (almost) solved the mathematical problem in question a couple of days ago. More on this later.

In any case, for those interested in the new breed of statistical measures designed to be truly outlier-resistant, noise-nonsensitive and size-independent (far more important requirements for big data, than ease of computation or mathematically-friendly) this old book might still be of interest.   

Views: 1558

Replies to This Discussion


I get $158.87 on my Amazon account. Still on the expensive side.  I see a used copy - Very Good for ~ $38.



On Data Science Central

© 2019 is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service