As the saying goes, Torture the data long enough, and it will confess to anything." With such a view, it is easy to see why I approach data mining with careful skepticism. Part of this skepticism comes from the fact that I work in the quantitative analytics group at an investment bank, so I often get to see some outrageous propositions for how to make a profit (justified by historical backtesting, of course). There is, however, no escaping the fascination with this field. Hence, despite any criticism I have, this is a great book to read. It is interesting, educational, and a pleasure to review. Each chapter lists and discusses available software implementations and ends with exercises for the reader, a bibliography of references for further reading, and a review of current research areas.
展开▼
