learning statistics and statistical learning are totally different, actually statistical learning is based on graduate level, which normally can be divided in 2 directions: supervised learning(e.g. BP-ANN, SVM) and unsupervised learning(e.g. hierarchical clustering)a very famous product based on this theory is the self-drived car developped by google(http://www.bbc.co.uk/news/technology-17989553). and two very valuable reference books, which could be seen as classics in this filed, 1,<Statistical Learning Theory> 2,<
The Nature of Statistical Learning Theory> both are written by the pioneer statistican Vladimir Vapnik. between 2 books, <
The Nature of Statistical Learning Theory> is actually easier for beginner, all proofs of this book can be found in <Statistical Learning Theory>, which is a pure theoretical reference.