Package com.datumbox.framework.applications.datamodeling

Class Modeler

java.lang.Object

com.datumbox.framework.core.machinelearning.common.abstracts.AbstractTrainer<MP,TP>

com.datumbox.framework.core.machinelearning.common.abstracts.modelers.AbstractModeler<Modeler.ModelParameters,Modeler.TrainingParameters>

com.datumbox.framework.applications.datamodeling.Modeler

All Implemented Interfaces:

Savable, Parallelizable, Trainable<Modeler.ModelParameters,Modeler.TrainingParameters>, AutoCloseable

Direct Known Subclasses:

TextClassifier

public class Modeler
extends AbstractModeler<Modeler.ModelParameters,Modeler.TrainingParameters>
implements Parallelizable

Modeler is a convenience class which can be used to train Machine Learning models. It is a wrapper class which automatically takes care of the data transformation, feature selection and modeler training processes.

Author:

Vasilis Vryniotis

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	Modeler.ModelParameters	It contains all the Model Parameters which are learned during the training.
static class	Modeler.TrainingParameters	It contains the Training Parameters of the Modeler.

Nested classes/interfaces inherited from class com.datumbox.framework.core.machinelearning.common.abstracts.AbstractTrainer

AbstractTrainer.AbstractModelParameters, AbstractTrainer.AbstractTrainingParameters

Field Summary

Fields

Modifier and Type	Field	Description
protected static String	CE_KEY	Key for Categorical Encoding.
protected static String	FS_KEY	Key for Feature Selection.
protected static String	ML_KEY	Key for Modeling.
protected static String	NS_KEY	Key for Numeric Scaling.
protected List<String>	pipeline	The steps of the pipeline.

Fields inherited from class com.datumbox.framework.core.machinelearning.common.abstracts.AbstractTrainer

knowledgeBase, logger

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	Modeler(Modeler.TrainingParameters trainingParameters, com.datumbox.framework.common.Configuration configuration)
protected	Modeler(String storageName, com.datumbox.framework.common.Configuration configuration)

Method Summary

Modifier and Type	Method	Description
protected void	_fit(Dataframe trainingData)
void	_predict(Dataframe newData)
void	close()
void	delete()
boolean	isParallelized()
void	save(String storageName)
void	setParallelized(boolean parallelized)

Methods inherited from class com.datumbox.framework.core.machinelearning.common.abstracts.modelers.AbstractModeler

predict

Methods inherited from class com.datumbox.framework.core.machinelearning.common.abstracts.AbstractTrainer

createKnowledgeBaseName, fit, getModelParameters, getTrainingParameters

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

NS_KEY

protected static final String NS_KEY

Key for Numeric Scaling.

See Also:

Constant Field Values

CE_KEY

protected static final String CE_KEY

Key for Categorical Encoding.

See Also:

Constant Field Values

FS_KEY

protected static final String FS_KEY

Key for Feature Selection.

See Also:

Constant Field Values

ML_KEY

protected static final String ML_KEY

Key for Modeling.

See Also:

Constant Field Values

pipeline

protected List<String> pipeline

The steps of the pipeline.

Constructor Detail

Modeler

protected Modeler(Modeler.TrainingParameters trainingParameters,
                  com.datumbox.framework.common.Configuration configuration)

Parameters:

trainingParameters -

configuration -

See Also:

AbstractTrainer(AbstractTrainingParameters, Configuration)

Modeler

protected Modeler(String storageName,
                  com.datumbox.framework.common.Configuration configuration)

Parameters:

storageName -

configuration -

See Also:

AbstractTrainer(java.lang.String, Configuration)

Method Detail

isParallelized

public boolean isParallelized()

Specified by:

isParallelized in interface Parallelizable

setParallelized

public void setParallelized(boolean parallelized)

Specified by:

setParallelized in interface Parallelizable

_predict

public void _predict(Dataframe newData)

Specified by:

_predict in class AbstractModeler<Modeler.ModelParameters,Modeler.TrainingParameters>

_fit

protected void _fit(Dataframe trainingData)

Specified by:

_fit in class AbstractTrainer<Modeler.ModelParameters,Modeler.TrainingParameters>

save

public void save(String storageName)

Specified by:

save in interface Savable

Overrides:

save in class AbstractTrainer<Modeler.ModelParameters,Modeler.TrainingParameters>

delete

public void delete()

Specified by:

delete in interface Savable

Overrides:

delete in class AbstractTrainer<Modeler.ModelParameters,Modeler.TrainingParameters>

close

public void close()

Specified by:

close in interface AutoCloseable

Overrides:

close in class AbstractTrainer<Modeler.ModelParameters,Modeler.TrainingParameters>