Introduction

Given the obvious importance of Modeling, why only one Chapter? Certainly, one could easily write 1000 pages on the various algorithms and the proof is in the large number of books that have done just that. The goal of this Cookbook, however, is to direct the reader to areas that they might otherwise spend too little time on, or to suggest approaches that are non-obvious.

There is much about the many algorithms that is non-obvious. They demand study. Thankfully, they also reward that study but in ways that can be frustrating to the intermediate-level data miner. It is often said, and can actually be shown, that detailed study of a handful of algorithms might be superior than spreading one's professional development time across all of them. It is worth noting that those that have the time and attention to learn R, which would also reward study, could learn hundreds of classifiers, and many hundreds of algorithms. The problem is that while mastering algorithms comes slowly...