public static class ISOUPTree.LeafNode extends ISOUPTree.Node
Modifier and Type | Field and Description |
---|---|
DoubleVector |
errorM |
DoubleVector |
errorP |
protected double |
examplesSeenAtLastSplitEvaluation |
ISOUPTree.MultitargetPerceptron |
learningModel |
alternateTree, attributeObservers, changeDetection, examplesSeen, ID, originalNode, parent, sumOfSquares, sumOfValues, tree, weightSeenAtLastSplitEvaluation
Constructor and Description |
---|
LeafNode(ISOUPTree tree)
Create a new LeafNode
|
Modifier and Type | Method and Description |
---|---|
void |
checkForSplit() |
void |
describeSubtree(StringBuilder out,
int indent) |
AttributeExpansionSuggestion[] |
getBestSplitSuggestions(MultiLabelSplitCriterion criterion)
Return the best split suggestions for this node using the given split criteria
|
double[] |
getPrediction(MultiLabelInstance inst) |
double[] |
getPredictionModel(MultiLabelInstance inst)
Retrieve the class votes using the perceptron learner
|
double[] |
getPredictionTargetMean(MultiLabelInstance inst) |
void |
learnFromInstance(MultiLabelInstance inst,
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
|
calcByteSize, copyStatistics, disableChangeDetection, getChildIndex, getDescription, getLevel, getNumSubtrees, getParent, restartChangeDetection, setChild, setParent, skipInLevelCount
copy, copy, measureByteSize, measureByteSize, toString
public ISOUPTree.MultitargetPerceptron learningModel
public DoubleVector errorP
public DoubleVector errorM
protected double examplesSeenAtLastSplitEvaluation
public LeafNode(ISOUPTree tree)
public void learnFromInstance(MultiLabelInstance inst, double[] prediction, boolean growthAllowed)
public AttributeExpansionSuggestion[] getBestSplitSuggestions(MultiLabelSplitCriterion criterion)
public double[] getPredictionModel(MultiLabelInstance inst)
public double[] getPredictionTargetMean(MultiLabelInstance inst)
public double[] getPrediction(MultiLabelInstance inst)
getPrediction
in class ISOUPTree.Node
public void checkForSplit()
public void describeSubtree(StringBuilder out, int indent)
describeSubtree
in class ISOUPTree.Node
Copyright © 2019 University of Waikato, Hamilton, NZ. All Rights Reserved.