SQL Server : How to Create OLAP Data Mining Models

The process of creating an OLAP data mining model is similar to the process of creating a relational data mining model. To create an OLAP mining model, use the Mining Model Wizard. The steps needed to create an OLAP mining model are nearly identical to those needed to create a relational mining model.

The Mining Model Wizard uses the source cube, data mining technique, case dimension and level, predicted entity, and training data to create an OLAP data mining model.

The source cube provides the Mining Model Wizard with the information needed to both create a case set for the data mining model. Because a cube may contain many groups of information, the model uses a dimension and level that you choose from the cube to establish key columns for the case set.

When you select the data mining technique, you also select a data mining provider. The data mining provider provides the data mining algorithms and model structure for the data mining model.

The case dimension and level provide a specific orientation for the data mining model into the cube for creating a case set.

The predicted entity can be one of the following entities:

A measure of the source cube


A member property of the case dimension and level, selected earlier in the Mining Model Wizard


Members of another dimension in the cube
This provides flexibility in dealing with the potentially complex process of predictive analysis on OLAP data.

Training data, in the form of dimensions, levels, member properties and measures, is used to process the OLAP data mining model and further define the data mining column structure for the case set.

Optionally, the wizard can create a new dimension for the source cube or a virtual cube based on the source cube. This enables users to query the data mining data just as they would query OLAP data.