Here is a bit more explanation from the introduction to the exercise document:
The third edition of Data Mining Techniques by Gordon S. Linoff and Michael J. A. Berry contains many stories about applying these techniques to real business data to achieve real business goals, but until now there has not been an easy way for students to follow along getting their hands dirty and learning from experience. This supplement fills that gap by providing suggested exercises along with the data to be mined.
We have always shied away from including exercises in our books for several reasons. One is that software evolves so rapidly that the exercises can quickly seem out-of-date. Another is that we wanted the freedom to talk about techniques and variations without worrying about whether they were supported by this or that software package. Finally, the foundations of data mining do not require learning a specific package and we did not want software to distract readers from the principal focus of the book. On the other hand, there is no substitute for hands-on work with real data for truly mastering data mining techniques. The solution is to publish a separate document with an update cycle decoupled from that of Data Mining Techniques.