moa.classifiers.trees

Class FIMTDD

java.lang.Object

moa.AbstractMOAObject

moa.options.AbstractOptionHandler

moa.classifiers.AbstractClassifier

moa.classifiers.trees.HoeffdingTree

moa.classifiers.trees.FIMTDD

All Implemented Interfaces:

Serializable, Classifier, Regressor, AWTRenderable, MOAObject, OptionHandler

public class FIMTDD
extends HoeffdingTree
implements Regressor

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static interface	FIMTDD.AdaptationCompatibleNode A new interface for nodes to be used in an adaptive setting
static class	FIMTDD.FIMTDDActiveLearningNode A modified ActiveLearningNode that uses a Perceptron as the leaf node model, and ensures that the class values sent to the attribute observers are not truncated to ints if regression is being performed
static class	FIMTDD.FIMTDDPerceptron A Perceptron classifier modified to conform to the specifications of Ikonomovska et al.
static class	FIMTDD.FIMTDDSplitNode A modified SplitNode method implementing the extra information regarding it's parent, and the ability to track the error rate and perform Page-Hinckley tests

Nested classes/interfaces inherited from class moa.classifiers.trees.HoeffdingTree

HoeffdingTree.ActiveLearningNode, HoeffdingTree.FoundNode, HoeffdingTree.InactiveLearningNode, HoeffdingTree.LearningNode, HoeffdingTree.LearningNodeNB, HoeffdingTree.LearningNodeNBAdaptive, HoeffdingTree.Node, HoeffdingTree.SplitNode

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	Adaptable
FloatOption	AlternateTreeFadingFactorOption
IntOption	AlternateTreeTimeOption
IntOption	AlternateTreeTMinOption
protected double	initLearnRate
FlagOption	learningRatio_Decay_or_Const_Option
FloatOption	learningRatioOption
protected double	learnRateDecay
protected ArrayList<FIMTDD.FIMTDDSplitNode>	nodesToAdapt
FloatOption	PageHinckleyAlphaOption
IntOption	PageHinckleyThresholdOption
protected DoubleVector	splitRatioStatistics

Fields inherited from class moa.classifiers.trees.HoeffdingTree

activeLeafByteSizeEstimate, activeLeafNodeCount, binarySplitsOption, byteSizeEstimateOverheadFraction, decisionNodeCount, gracePeriodOption, growthAllowed, inactiveLeafByteSizeEstimate, inactiveLeafNodeCount, leafpredictionOption, maxByteSizeOption, memoryEstimatePeriodOption, nbThresholdOption, nominalEstimatorOption, noPrePruneOption, numericEstimatorOption, removePoorAttsOption, splitConfidenceOption, splitCriterionOption, stopMemManagementOption, tieThresholdOption, treeRoot

Fields inherited from class moa.classifiers.AbstractClassifier

classifierRandom, modelContext, randomSeed, randomSeedOption, trainingWeightSeenByModel

Fields inherited from class moa.options.AbstractOptionHandler

classOptionNamesToPreparedObjects, options

Constructor Summary

Constructors

Constructor and Description
FIMTDD()

Method Summary

Methods

Modifier and Type	Method and Description
protected void	FIMTDD_attemptToSplit(FIMTDD.FIMTDDActiveLearningNode node, FIMTDD.FIMTDDSplitNode parent, int parentIndex) Method used to split a leaf node and generate child nodes, if appropriate
String	getPurposeString() Gets the purpose of this object
protected FIMTDD.FIMTDDActiveLearningNode	newLearningNode() Return an empty FIMTDDActiveLearningNode
protected FIMTDD.FIMTDDActiveLearningNode	newLearningNode(double[] initialClassObservations) Return a new FIMTDDActiveLearningNode using the initial class observations
protected FIMTDD.FIMTDDSplitNode	newSplitNode(InstanceConditionalTest splitTest, double[] classObservations) Return a new FIMTDDSplitNode
void	trainOnInstanceImpl(weka.core.Instance inst) Method for updating (training) the model using a new instance

Methods inherited from class moa.classifiers.trees.HoeffdingTree

activateLearningNode, attemptToSplit, calcByteSize, computeHoeffdingBound, deactivateAllLeaves, deactivateLearningNode, enforceTrackerLimit, estimateModelByteSizes, findLearningNodes, findLearningNodes, getModelDescription, getModelMeasurementsImpl, getVotesForInstance, isRandomizable, measureByteSize, measureTreeDepth, newNominalClassObserver, newNumericClassObserver, newSplitNode, resetLearningImpl

Methods inherited from class moa.classifiers.AbstractClassifier

contextIsCompatible, copy, correctlyClassifies, getAttributeNameString, getAWTRenderer, getClassLabelString, getClassNameString, getDescription, getModelContext, getModelMeasurements, getNominalValueString, getSubClassifiers, modelAttIndexToInstanceAttIndex, modelAttIndexToInstanceAttIndex, prepareForUseImpl, resetLearning, setModelContext, setRandomSeed, trainingHasStarted, trainingWeightSeenByModel, trainOnInstance

Methods inherited from class moa.options.AbstractOptionHandler

discoverOptionsViaReflection, getCLICreationString, getOptions, getPreparedClassOption, getPreparedClassOption, prepareClassOptions, prepareForUse, prepareForUse

Methods inherited from class moa.AbstractMOAObject

copy, measureByteSize, toString

Methods inherited from class java.lang.Object

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

Methods inherited from interface moa.options.OptionHandler

getCLICreationString, getOptions, prepareForUse, prepareForUse

Field Detail

splitRatioStatistics

protected DoubleVector splitRatioStatistics

nodesToAdapt

protected ArrayList<FIMTDD.FIMTDDSplitNode> nodesToAdapt

Adaptable

protected boolean Adaptable

initLearnRate

protected double initLearnRate

learnRateDecay

protected double learnRateDecay

PageHinckleyAlphaOption

public FloatOption PageHinckleyAlphaOption

PageHinckleyThresholdOption

public IntOption PageHinckleyThresholdOption

AlternateTreeFadingFactorOption

public FloatOption AlternateTreeFadingFactorOption

AlternateTreeTMinOption

public IntOption AlternateTreeTMinOption

AlternateTreeTimeOption

public IntOption AlternateTreeTimeOption

learningRatioOption

public FloatOption learningRatioOption

learningRatio_Decay_or_Const_Option

public FlagOption learningRatio_Decay_or_Const_Option

Constructor Detail

FIMTDD

public FIMTDD()

Method Detail

getPurposeString

public String getPurposeString()

Description copied from interface: OptionHandler

Gets the purpose of this object

Specified by:

getPurposeString in interface OptionHandler

Overrides:

getPurposeString in class HoeffdingTree

Returns:

the string with the purpose of this object

trainOnInstanceImpl

public void trainOnInstanceImpl(weka.core.Instance inst)

Method for updating (training) the model using a new instance

Overrides:

trainOnInstanceImpl in class HoeffdingTree

Parameters:

inst - the instance to be used for training

FIMTDD_attemptToSplit

protected void FIMTDD_attemptToSplit(FIMTDD.FIMTDDActiveLearningNode node,
                         FIMTDD.FIMTDDSplitNode parent,
                         int parentIndex)

Method used to split a leaf node and generate child nodes, if appropriate

newLearningNode

protected FIMTDD.FIMTDDActiveLearningNode newLearningNode()

Return an empty FIMTDDActiveLearningNode

Overrides:

newLearningNode in class HoeffdingTree

newLearningNode

protected FIMTDD.FIMTDDActiveLearningNode newLearningNode(double[] initialClassObservations)

Return a new FIMTDDActiveLearningNode using the initial class observations

Overrides:

newLearningNode in class HoeffdingTree

newSplitNode

protected FIMTDD.FIMTDDSplitNode newSplitNode(InstanceConditionalTest splitTest,
                                  double[] classObservations)

Return a new FIMTDDSplitNode

Overrides:

newSplitNode in class HoeffdingTree