org.encog.ensemble

Class Ensemble

java.lang.Object

org.encog.ensemble.Ensemble

Direct Known Subclasses:

AdaBoost, Bagging, Stacking

public abstract class Ensemble
extends Object

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	Ensemble.NotPossibleInThisMethod

Field Summary

Fields

Modifier and Type	Field and Description
protected EnsembleAggregator	aggregator
protected MLDataSet	aggregatorDataSet
protected EnsembleDataSetFactory	dataSetFactory
protected ArrayList<EnsembleML>	members
protected EnsembleMLMethodFactory	mlFactory
protected EnsembleTrainFactory	trainFactory

Constructor Summary

Constructors

Constructor and Description
Ensemble()

Method Summary

Methods

Modifier and Type	Method and Description
void	addMember(EnsembleML newMember) Add a member to the ensemble
MLData	compute(MLData input) Compute the output for a specific input
EnsembleAggregator	getAggregator()
EnsembleML	getMember(int memberNumber) Extract a specific MLMethod
abstract EnsembleTypes.ProblemType	getProblemType() Return what type of problem this Ensemble is solving
MLDataSet	getTrainingSet(int setNumber) Extract a specific training set from the Ensemble
abstract void	initMembers() Initialise ensemble components
void	initMembersBySplits(int splits)
void	setAggregator(EnsembleAggregator aggregator) Sets the ensemble aggregation method
void	setTrainingData(MLDataSet data) Set which training data to base the training on
void	setTrainingDataFactory(EnsembleDataSetFactory dataSetFactory) Set which dataSetFactory to use to create the correct tranining sets
void	setTrainingMethod(EnsembleTrainFactory newTrainFactory) Set the training method to use for this ensemble
void	train(double targetError, double selectionError, EnsembleDataSet testset) Train the ensemble to a target accuracy
void	train(double targetError, double selectionError, EnsembleDataSet testset, boolean verbose) Train the ensemble to a target accuracy

Methods inherited from class java.lang.Object

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

Field Detail

dataSetFactory

protected EnsembleDataSetFactory dataSetFactory

trainFactory

protected EnsembleTrainFactory trainFactory

aggregator

protected EnsembleAggregator aggregator

members

protected ArrayList<EnsembleML> members

mlFactory

protected EnsembleMLMethodFactory mlFactory

aggregatorDataSet

protected MLDataSet aggregatorDataSet

Constructor Detail

Ensemble

public Ensemble()

Method Detail

initMembers

public abstract void initMembers()

Initialise ensemble components

initMembersBySplits

public void initMembersBySplits(int splits)

setTrainingMethod

public void setTrainingMethod(EnsembleTrainFactory newTrainFactory)

Set the training method to use for this ensemble

Parameters:

newTrainFactory - The training factory.

setTrainingData

public void setTrainingData(MLDataSet data)

Set which training data to base the training on

Parameters:

data - The training data.

setTrainingDataFactory

public void setTrainingDataFactory(EnsembleDataSetFactory dataSetFactory)

Set which dataSetFactory to use to create the correct tranining sets

Parameters:

dataSetFactory - The data set factory.

train

public void train(double targetError,
         double selectionError,
         EnsembleDataSet testset,
         boolean verbose)

Train the ensemble to a target accuracy

Parameters:

targetError - The target error.

selectionError - The selection error.

testset - The test set.

verbose - Verbose mode?

train

public void train(double targetError,
         double selectionError,
         EnsembleDataSet testset)

Train the ensemble to a target accuracy

Parameters:

targetError - The target error.

selectionError - The selection error.

testset - The test set.

getTrainingSet

public MLDataSet getTrainingSet(int setNumber)

Extract a specific training set from the Ensemble

Parameters:

setNumber -

Returns:

The training set.

getMember

public EnsembleML getMember(int memberNumber)

Extract a specific MLMethod

Parameters:

memberNumber -

Returns:

The MLMethod.

addMember

public void addMember(EnsembleML newMember)
               throws Ensemble.NotPossibleInThisMethod

Add a member to the ensemble

Parameters:

newMember -

Throws:

Ensemble.NotPossibleInThisMethod

compute

public MLData compute(MLData input)

Compute the output for a specific input

Parameters:

input -

Returns:

The data.

getAggregator

public EnsembleAggregator getAggregator()

Returns:

Returns the ensemble aggregation method

setAggregator

public void setAggregator(EnsembleAggregator aggregator)

Sets the ensemble aggregation method

Parameters:

aggregator -

getProblemType

public abstract EnsembleTypes.ProblemType getProblemType()

Return what type of problem this Ensemble is solving

Returns:

The problem type.