The most common basic data mining tasks (supported by SQL Server 2005) are classification, regression, clustering, forecasting, association analysis, correlation analysis.
All tasks act as blocks in solving a really large set of business problems.
A few starting points:
- http://www.microsoft.com/sql/technologies/dm/default.mspx -- The tutorial and the Webcast ("Introduction to Data Mining in SQL Server 2005") provide a good description of what data mining tasks are supported in SQL Server, plus details on the tools
- http://www.sqlserverdatamining.com -- lots of information, tutorials etc.
|||
A nice classification of data mining tasks is given by David Hand et.al in Principles of Data Mining (2001):
1. Exploratory data analysis (Visualization)
2. Descriptive modelling (Cluster analysis, relationships between variables...)
3. Predictive Modelling (Regression, Classification,...)
4. Pattern Discovery (Association Rules, outlier detection,...)
5. Retrieval by content (start from a given pattern and try to find similar patterns)
This provides a very general viewpoint on all different kinds of tasks.
