moa.clusterers.clustree

Class ClusTree

java.lang.Object

moa.AbstractMOAObject

moa.options.AbstractOptionHandler

moa.clusterers.AbstractClusterer

moa.clusterers.clustree.ClusTree

All Implemented Interfaces:

Serializable, Clusterer, AWTRenderable, MOAObject, OptionHandler

Direct Known Subclasses:

AnyOutCore

public class ClusTree
extends AbstractClusterer

Citation: ClusTree: Philipp Kranen, Ira Assent, Corinna Baldauf, Thomas Seidl: The ClusTree: indexing micro-clusters for anytime stream mining. Knowl. Inf. Syst. 29(2): 249-272 (2011)

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	breadthFirstStrat Parameter to determine wich strategy to use
IntOption	horizonOption
protected int	maxHeight The maximal height of the tree.
IntOption	maxHeightOption
protected double	negLambda Parameter for the weighting function use to weight the entries.
protected Node	root The root node of the tree.

Fields inherited from class moa.clusterers.AbstractClusterer

clustererRandom, clustering, evaluateMicroClusteringOption, modelContext, randomSeed, randomSeedOption, trainingWeightSeenByModel

Fields inherited from class moa.options.AbstractOptionHandler

classOptionNamesToPreparedObjects, options

Constructor Summary

Constructors

Constructor and Description
ClusTree()

Method Summary

Methods

Modifier and Type	Method and Description
Clustering	getClustering(long currentTime, int targetLevel)
Clustering	getClusteringResult()
protected int	getDefaultHeight()
int	getHeight() Return the current height of the tree.
Clustering	getMicroClusteringResult()
void	getModelDescription(StringBuilder out, int indent)
protected Measurement[]	getModelMeasurementsImpl()
int	getNumRootSplits() Return the number of time the tree has grown in size.
double[]	getVotesForInstance(weka.core.Instance inst)
boolean	implementsMicroClusterer()
void	insert(ClusKernel newPoint, Budget budget, long timestamp) Insert a new point in the Tree.
boolean	isRandomizable()
void	resetLearningImpl()
void	trainOnInstanceImpl(weka.core.Instance instance)

Methods inherited from class moa.clusterers.AbstractClusterer

contextIsCompatible, copy, getAttributeNameString, getAWTRenderer, getClassLabelString, getClassNameString, getDescription, getModelContext, getModelMeasurements, getNominalValueString, getPurposeString, getSubClusterers, keepClassLabel, 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, 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

Methods inherited from interface moa.MOAObject

measureByteSize

Field Detail

horizonOption

public IntOption horizonOption

maxHeightOption

public IntOption maxHeightOption

root

protected Node root

The root node of the tree.

negLambda

protected double negLambda

Parameter for the weighting function use to weight the entries.

maxHeight

protected int maxHeight

The maximal height of the tree.

breadthFirstStrat

protected boolean breadthFirstStrat

Parameter to determine wich strategy to use

Constructor Detail

ClusTree

public ClusTree()

Method Detail

getDefaultHeight

protected int getDefaultHeight()

resetLearningImpl

public void resetLearningImpl()

Specified by:

resetLearningImpl in class AbstractClusterer

getModelMeasurementsImpl

protected Measurement[] getModelMeasurementsImpl()

Specified by:

getModelMeasurementsImpl in class AbstractClusterer

isRandomizable

public boolean isRandomizable()

getModelDescription

public void getModelDescription(StringBuilder out,
                       int indent)

Specified by:

getModelDescription in class AbstractClusterer

getVotesForInstance

public double[] getVotesForInstance(weka.core.Instance inst)

implementsMicroClusterer

public boolean implementsMicroClusterer()

Specified by:

implementsMicroClusterer in interface Clusterer

Overrides:

implementsMicroClusterer in class AbstractClusterer

trainOnInstanceImpl

public void trainOnInstanceImpl(weka.core.Instance instance)

Specified by:

trainOnInstanceImpl in class AbstractClusterer

insert

public void insert(ClusKernel newPoint,
          Budget budget,
          long timestamp)

Insert a new point in the Tree. The point should be represented as a cluster with a single data point(i.e. N = 1). A Budget class is also given, which is informed of the number of operation the tree does, and informs the tree when it does not have time left and should stop the insertion.

Parameters:

newPoint - The point to be inserted.

budget - The budget and statistics recollector for the insertion.

timestamp - The moment at which this point is inserted.

See Also:

Kernel, Budget

getNumRootSplits

public int getNumRootSplits()

Return the number of time the tree has grown in size. If the tree grows and is then cutted from a certain depth, it also counts.

Returns:

The number of times the root node was splitted.

getHeight

public int getHeight()

Return the current height of the tree. This should never be greater than maxHeight.

Returns:

The height of the tree.

See Also:

maxHeight

getMicroClusteringResult

public Clustering getMicroClusteringResult()

Specified by:

getMicroClusteringResult in interface Clusterer

Overrides:

getMicroClusteringResult in class AbstractClusterer

Parameters:

currentTime - The current time

Returns:

The kernels at the leaf level as a clustering

getClusteringResult

public Clustering getClusteringResult()

getClustering

public Clustering getClustering(long currentTime,
                       int targetLevel)

Parameters:

currentTime - The current time

Returns:

The kernels at the given level as a clustering.