Good book for learning about the data mining techniques of logistic and linear regression. It helped highlight some good uses, and fortunately, I've recently had the opportunity to use it in my work.

However, I was a bit disappointed that the data preparation seemed very coding intensive. The author could have shown readers how to merge lookup tables of risk values onto customer datasets, rather than hard-coding each of the rules and values; or to use the SAS procedure for creating indicator variables, instead of writing the rules for each category.

Overall, I'm glad that I purchased the book - it lives up to its claims - but it misses some of the better practices, and time saving devices, in data preparation

Скачать книгу бесплатно (pdf, 6.75 Mb)

---

Читать «Data Mining Cookbook: Modeling Data for Marketing, Risk and Customer Relationship Management»