Adaptive Model form Completing the Configuration tab
About  New Configuration Settings Pages & Classes History

Use this tab to configure to define the behavior of the adaptive model.

Properties in the data model can be used as predictors. If an adaptive model does not have explicitly defined predictors, ADM dynamically adjusts to keep all data encountered within its internal representation of the data. You should select all properties that have potential predictive performance. The adaptive analytics engine automatically detects the most important predictive fields to use in an adaptive scoring model.

The capability of using predictors depends is added by the Entitled to Use Adaptive Decision Manager setting in the Services landing page. If this setting is disabled, you can define and use adaptive models in your application, but these models can not operate based on predictive data.

The models in ADM that are configured by this rule learn from the behavior defined in the Positive Behavior and Negative Behavior sections of this tab. The values that are displayed correspond to the possible behavior defined in the system for the behavior dimension, which consists of the combination provided for type of behavior (typically, positive, or negative), and response (for example, Accept, and Reject). ADM can also learn from response values in the interaction result that are not defined in these sections. In this case, you simply define the behavior type in the corresponding behavior section (for example, Positive in the Positive Behavior sections, and Negative in the Negative Behavior section), and ADM learns from responses set as, for example, Positive-Accept, Positive-Yes, Negative-Reject, Negative-No, etc.

Field	Description
Predictors	Use this section to add as many rows as predictors you want to use in the adaptive scoring model. Predictors can be treated as numeric or symbolic. Click to add a row or to delete a row.
Property	Select an existing property or click to create a new property and map it.
Property Type	Defaults to the data type of property selected.
Predictor Type	Select the predictor data type for the property. Predictors can be treated as numeric or symbolic.
Positive Behavior	Use this section to map a list of values you want to associate with positive behavior.
Negative Behavior	Use this section to map a list of values you want to associate with negative behavior.

About Adaptive Model rules