This section describes setting the target field, maximum
tolerated systematic errors and active or inactive fields for Regression
mining function.
To specify settings for the Regression mining function, you can
specify the following parameters:
- Target field
- The specification of the target
field is mandatory.
To specify the field name of a target field
of a DM_RegSettings object, you can use DM_setRegTarget.
The
mining data type of the target field must be numerical. The name of
the target field must coincide with one of the field names that are
defined in the logical data specifications.
- Maximum tolerated systematic error (RSquared)
- To set an upper limit
for the acceptable systematic error of the model, you must specify
an upper threshold for r². To do this, you can use DM_setRegRSquared.
- Active or inactive fields
- The Regression
mining function detects fields that are not useful for a Regression
training run. It automatically sets these fields to inactive.
Inactive fields are ignored during the calculation of a Regression
model. The other fields are considered active and are therefore
used for the Regression mining run.
You can overrule the automatic
selection of active and inactive fields by using DM_setFldUsageType.