Package moa.classifiers.multilabel.trees
Class ISOUPTree.LeafNode
- java.lang.Object
-
- moa.classifiers.multilabel.trees.ISOUPTree.Node
-
- moa.classifiers.multilabel.trees.ISOUPTree.LeafNode
-
- All Implemented Interfaces:
Serializable
- Enclosing class:
- ISOUPTree
public static class ISOUPTree.LeafNode extends ISOUPTree.Node
- See Also:
- Serialized Form
-
-
Field Summary
Fields Modifier and Type Field Description protected AutoExpandVector<AttributeStatisticsObserver>
attributeObservers
DoubleVector
errorM
DoubleVector
errorP
protected double
examplesSeenAtLastSplitEvaluation
List<Integer>
inputIndexes
ISOUPTree.MultitargetPerceptron
learningModel
double
learningWeight
-
Fields inherited from class moa.classifiers.multilabel.trees.ISOUPTree.Node
alternateTree, changeDetection, examplesSeen, ID, originalNode, parent, sumOfInputSquares, sumOfInputValues, sumOfSquares, sumOfValues, tree, weightOfInputs, weightSeenAtLastSplitEvaluation
-
-
Method Summary
All Methods Instance Methods Concrete Methods Modifier and Type Method Description long
calcByteSize()
void
checkForSplit()
void
describeSubtree(StringBuilder out, int indent)
AttributeExpansionSuggestion[]
getBestSplitSuggestions(MultiLabelSplitCriterion criterion)
Return the best split suggestions for this node using the given split criteriadouble[]
getPrediction(Instance inst)
double[]
getPredictionModel(Instance inst)
Retrieve the class votes using the perceptron learnerdouble[]
getPredictionTargetMean(Instance inst)
void
initializeInputIndexes()
void
learnFromInstance(Instance instance, double[] prediction, boolean growthAllowed)
Method to learn from an instance that passes the new instance to the perceptron learner, and also prevents the class value from being truncated to an int when it is passed to the attribute observer-
Methods inherited from class moa.classifiers.multilabel.trees.ISOUPTree.Node
copyStatistics, disableChangeDetection, getChildIndex, getDescription, getLevel, getNumSubtrees, getParent, restartChangeDetection, setChild, setParent, skipInLevelCount
-
-
-
-
Field Detail
-
learningModel
public ISOUPTree.MultitargetPerceptron learningModel
-
learningWeight
public double learningWeight
-
errorP
public DoubleVector errorP
-
errorM
public DoubleVector errorM
-
examplesSeenAtLastSplitEvaluation
protected double examplesSeenAtLastSplitEvaluation
-
attributeObservers
protected AutoExpandVector<AttributeStatisticsObserver> attributeObservers
-
-
Constructor Detail
-
LeafNode
public LeafNode(ISOUPTree tree)
Create a new LeafNode
-
-
Method Detail
-
calcByteSize
public long calcByteSize()
- Overrides:
calcByteSize
in classISOUPTree.Node
-
initializeInputIndexes
public void initializeInputIndexes()
-
learnFromInstance
public void learnFromInstance(Instance instance, double[] prediction, boolean growthAllowed)
Method to learn from an instance that passes the new instance to the perceptron learner, and also prevents the class value from being truncated to an int when it is passed to the attribute observer
-
getBestSplitSuggestions
public AttributeExpansionSuggestion[] getBestSplitSuggestions(MultiLabelSplitCriterion criterion)
Return the best split suggestions for this node using the given split criteria
-
getPredictionModel
public double[] getPredictionModel(Instance inst)
Retrieve the class votes using the perceptron learner
-
getPredictionTargetMean
public double[] getPredictionTargetMean(Instance inst)
-
getPrediction
public double[] getPrediction(Instance inst)
- Specified by:
getPrediction
in classISOUPTree.Node
-
checkForSplit
public void checkForSplit()
-
describeSubtree
public void describeSubtree(StringBuilder out, int indent)
- Overrides:
describeSubtree
in classISOUPTree.Node
-
-