The Microsoft® Data Warehouse Toolkit: With SQL Server™ 2005 and the Microsoft® Business Intelligence Toolset

10.6. Summary

Congratulations for making it this far. You should have a basic understanding of the data mining concepts and how data mining can have an impact on your organization. Data mining is not just a tool; it's a process of understanding business needs and data issues, working through various alternative models, testing and validating their viability, rolling them out into production, and making sure they address the opportunity and don't get stale.

Early in the chapter we reviewed some basic data mining concepts, describing how data mining is used for several different business tasks: classification, estimation or regression, prediction, association or affinity grouping, clustering or segmentation, anomaly detection, and description and profiling. We then discussed SQL Server 2005's data mining architecture and toolset, reviewing the key components and showing how they fit together. Digging down into the technology, we briefly described the seven algorithms provided with the product and how they applied to the various business tasks.

Next we went into some detail on the process of data mining, outlining a step-by-step approach starting with identifying business opportunities and the associated data, moving through the actual data mining phase with its data preparation, model development, and model validation steps, and ending with the operations phase with the implementation of the model, maintenance, and an assessment of its impact.

Most of the second part of the chapter ...