Mining Model Wizard

Use this wizard to create a data mining model. A mining model enables you to analyze your data for patterns and to make predictions based on the patterns. You can create a mining model from a relational schema or a cube, and you can store output from the model in a tabular column, a cube dimension, or a mining model diagram.

The Mining Model Wizard appears when you perform any of the following actions:

In the Analysis Manager tree pane, right-click the Mining Models folder, and then click New Mining Model.
In the Analysis Manager tree pane, right-click a cube, and then click New Mining Model.
In the Analysis Manager tree pane, select a cube, and then on the Action menu, click New Mining Model.

After the Introduction step, the Mining Model Wizard begins with the following step if you start the wizard by right-clicking the Mining Models folder or by clicking New Mining Model on the Action menu:

Choose the type of data source for the mining model. You can create it from:
- Relational tables. This option creates a relational mining model.
- A cube. This option creates a multidimensional mining model.

The next steps depend on the type of data source you select. For a relational mining model, the Mining Model Wizard has the following steps:

Select the table or tables for your mining model.
Select the data mining technique to be used by your mining model.
Edit joins. This step is displayed only if you are creating a mining model from multiple tables.
Select the case key column for your mining model.
Select the input and predictable columns. You select predictable columns only if you select Microsoft Decision Trees as your data mining technique.
Finish. Name and save your mining model and optionally process it to view its results.

For OLAP data sources, which are directly specified either by using the wizard or by starting the wizard from selecting a cube, the Mining Model Wizard has the following steps:

Select the source cube for your mining model. If you began the wizard by right-clicking a cube, this step is not displayed.
Select the data mining technique to be used by your mining model.
Select the case dimension and level to be analyzed by the mining model.
Select the initial predicted entity. This can be a measure of the source cube, a member property of the case level, or members of another dimension. This step appears only if you select Microsoft Decision Trees as your data mining technique.
Select training data. Training data is known data that is represented by cube elements such as dimensions, levels, member properties, or measures. These are used by the model, the case dimension and the case level to derive the predicted entity by extrapolating from the known training data.
(Optional.) Create a dimension and/or a virtual cube. The wizard can create a dimension and a virtual cube from the results of the model analysis, enabling you to browse the model, include its cubes, or compare it to the source data. This step appears if you select Microsoft Decision Trees as your data mining technique.
Finish. Name and save your mining model and optionally process it to view its results.

Mining Model Wizard

See Also