This section describes various methods that
let you include or exclude fields in a model by declaring fields as
active or inactive.
Intelligent Miner® includes
a module for automatic variable selection. This module determines
the usage type of input fields. If the module determines a field not
to be useful for the mining process, the field is considered inactive
during the generation of the mining model.
For classification and regression,
you can let the automatic variable selection determine the active
fields and the inactive fields in your input data. You can also control
the automatic variable selection by specifying a usage type for a
field.
You can define the fields that you want to include in a model by
calling the method
DM_setFldUsageType() on this field
and by specifying one of the following usage types:
- Active
- To use the fields to build the model.
- Inactive
- To ignore the fields when an Associations model, a Classification
model, a Sequence Rules model, or a Regression model is built.
Note: For
Clustering models, the usage type is called supplementary.
With the following exceptions, you can also use the method
DM_setFldUsageType to
set all fields at once to a certain usage type:
- The target field for Classification and Regression
- The sequence field for Sequence rules, and the group field and
the item field for Association rules and Sequence rules
These fields are explicitly selected for a special purpose. You
need not call the method
DM_setFldUsageType on these
fields.
For more information refer to DM_setFldUsageType and Automatic variable selection.