Knight's Microsoft® Business Intelligence 24-Hour Trainer: Leveraging Microsoft SQL Server® Integration, Analysis, and Reporting Services with Excel® and SharePoint®

Data Mining

Data mining is one of the most exciting technologies in the Microsoft BI stack. Did you ever

wonder how Amazon.com can recommend such good books for you and generally be right? In

this lesson, you’ll learn how to use data mining to make predictions, explore your data, and

find patterns among your customers that you may not have been aware of.

INTRODUCTION TO DATA MINING

Data mining has been around for decades, but what Microsoft has revolutionized are its ease

of use and development. Data mining can solve more grand problems than most technologies

at a fairly minimal cost. SQL Server data mining ships with nine algorithms, and more can be

added if you’re feeling adventurous.

The first algorithm to highlight is the most common: the Decision Tree Algorithm. You can

use this if your management wants to know who is classified as a good customer as opposed to

a bad customer. While some old-timers in the organization may know this, it’s very subjective.

Data mining simplifies and automates that classification through the decision tree algorithm.

The most practical application for this algorithm is to answer the question “Will this customer

buy a product from me if I mail him a brochure?”

Data mining can also find fraud in insurance claims or applications for credit. In this case

the Clustering Algorithm measures the insurance claim’s distance from similar cases to see

whether any