gov.sandia.cognition.framework.learning
Interface CognitiveModuleFactoryLearner

All Superinterfaces:

Cloneable, gov.sandia.cognition.util.CloneableSerializable, Serializable

All Known Implementing Classes:

EvaluatorBasedCognitiveModuleFactoryLearner

public interface CognitiveModuleFactoryLearner
extends gov.sandia.cognition.util.CloneableSerializable

The CognitiveModuleFactoryLearner is an interface defining the functionality of an Object that can learn a CognitiveModuleFactory from a collection of input data.

Since:

2.0

Author:

Justin Basilico, Kevin R. Dixon

Method Summary

CognitiveModuleFactory

learn(CognitiveModel model, Collection<? extends Collection<? extends CognitiveModelInput>> datasets) Learns a new CognitiveModuleFactory for the given CognitiveModuleFactory containing all of the modules that will be used before the created module factory along with the example data used to learn the factory from.

Methods inherited from interface gov.sandia.cognition.util.CloneableSerializable

clone

Method Detail

learn

CognitiveModuleFactory learn(CognitiveModel model,
                             Collection<? extends Collection<? extends CognitiveModelInput>> datasets)

Learns a new CognitiveModuleFactory for the given CognitiveModuleFactory containing all of the modules that will be used before the created module factory along with the example data used to learn the factory from.

Parameters:

model - CognitiveModel to learn the new CognitiveModuleFactory for. Note that this has to be a CognitiveModel, not a CognitiveModelFactory, as the CognitiveModelInputs (needed for the dataset) needs SemanticIdentifier, which is specific to a CognitiveModel. Since the dataset must be created before this method call, the CognitiveModel used to give the SemanticIdentifiers to the dataset must also be the parameter here.

datasets - The datasets containing the inputs to the CognitiveModel for one or more sequences of data

Returns:

The CognitiveModuleFactory learned based on the given model factory and input data.