Craig Borysowich has some useful posts on his blog this week. He does nice little series on various topics and he recently completed a data mining series. Here are the links with some comments:
- Obtaining Management Support for Data Mining and BI
I blogged in response to this one already here
- Preparing Data for Data Mining
Craig correctly points out that this is probably the most difficult and most important step. As he says, it is critical to know what business problem you are trying to solve, what decision you are trying to influence, before starting. He also notes the importance of considering business rules as part of this process.
- Mining Data & Reviewing Results
This one is a little focused on data mining at the expense of predictive analytics but still interesting. I would recommend this post on the difference between predictive reporting and decision management as well as this one about the differences between BI, data mining and predictive analytics.
- Applying Data Mining Results
Short but a nice focus on the need for action as a result of data mining - those who act (right) first, win.
- Better Data Mining & BI
Note sure about this one - not 100% clear what Craig is getting at. On books I would recommend a couple of others too: Tom Davenport's "Competing on Analytics" (reviewed here) Berry and Linoff's classic "Data Mining Techniques" (reviewed here) ... and, of course, my new book "Smart (Enough) Systems" :-)
Technorati Tags: analytic application, analytics, business rules, competing on analytics, data mining, decision analysis, decision management, predictive analytics, Smart (Enough) Systems, smartenoughsystems