Package com.tibco.patterns.learn.modelconfig

Configuration and options for creating and using Learn models.

The ModelConfig class provides methods to store valid model configuration parameters and to create models based on these parameters. The SubsetTrainConfig abstract class and its ST... subclasses hold valid configuration parameters for each subset training mode. They are used together with the ModelConfig class.

The ConfQ... classes extend confidence measure classes to provide means to configure a Patterns query to use the specific confidence measure.

Interface Summary

Interface	Description
ConfQuery	Defines a confidence measure that can be used in a Patterns server RLink query.

Class Summary

Class	Description
ConfQFeature	Extends the ConfFeature class to provide means to configure a NetricsQuery.
ConfQNone	Extends the ConfNone class to provide means to configure a NetricsQuery.
ModelConfig	Stores a configuration of an RLink model and builds model(s) with this configuration.
STAll	Specifies training of all subsets with generated examples.
STDynamic	Specifies dynamic training of subsets.
STFixed	Specifies training up to a fixed depth of subsets, defined by percentage of missing information in the subset.
STNone	Specifies training with actual examples only, without any generated examples for submodels.
STTaper	Specifies the use of a random tapering function to generate some examples for subsets.
SubsetTrainConfig	Defines configuration parameters for one of the subset training modes.