So, this is a actually a draft, which is kindly kept available online by the author (see here). Even though it's a draft, I believe it deserves a rating already. One can't simply spoil something that is already too good unless you get to publish the final version entirely in Comic Sans font. Matloff's work already feels solid enough to be recommended wherever machine learning is taught. Bonus points if machine learning is approached via R, though the clean and straightforward example codes are far from restrictive. In this book, Matloff teaches you machine learning through real-world examples and very pertinent thought experiments. I don't know how the author does that, but he pinpointed all the 'what-if's that every machine learning enthusiast has certainly come up with at some point.