keel.Algorithms.Semi_Supervised_Learning.CoForest

Class RandomTree

java.lang.Object

keel.Algorithms.Semi_Supervised_Learning.CoForest.RandomTree

public class RandomTree
extends java.lang.Object

Class for constructing a tree that considers K randomly chosen attributes at each node. Performs no pruning. Also has an option to allow estimation of class probabilities based on a hold-out set (backfitting).

Valid options are:

 -K <number of attributes>
  Number of attributes to randomly investigate
  (<0 = int(log_2(#attributes)+1)).

 -M <minimum number of instances>
  Set minimum number of instances per leaf.

 -S <num>
  Seed for random number generator.
  (default 1)

 -depth <num>
  The maximum depth of the tree, 0 for unlimited.
  (default 0)

 -N <num>
  Number of folds for backfitting (default 0, no backfitting).

 -U
  Allow unclassified instances.

 -D
  If set, classifier is run in debug mode and
  may output additional info to the console

Version:

$Revision: 5928 $

Author:

Eibe Frank (eibe@cs.waikato.ac.nz), Richard Kirkby (rkirkby@cs.waikato.ac.nz)

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	m_AllowUnclassifiedInstances Whether unclassified instances are allowed
protected int	m_Attribute The attribute to split on.
protected double[]	m_ClassDistribution Class probabilities from the training data.
protected boolean	m_Debug Debug flag.
protected PrototypeSet	m_Info The header information.
protected int	m_KValue The number of attributes considered for a split.
protected int	m_MaxDepth The maximum depth of the tree (0 = unlimited)
protected double	m_MinNum Minimum number of instances for leaf.
protected int	m_NumFolds Determines how much data is used for backfitting
protected double[]	m_Prop The proportions of training instances going down each branch.
protected int	m_randomSeed The random seed to use.
protected double	m_SplitPoint The split point.
protected RandomTree[]	m_Successors The subtrees appended to this tree.
protected int	nClasses Number of classes.

Constructor Summary

Constructors

Constructor and Description
RandomTree()

Method Summary

Modifier and Type	Method and Description
java.lang.String	allowUnclassifiedInstancesTipText() Returns the tip text for this property
void	backfitData(PrototypeSet data) Backfits the given data into the tree.
protected void	backfitData(PrototypeSet data, double[] classProbs) Recursively backfits data into the tree.
void	buildClassifier(PrototypeSet data) Builds classifier.
protected void	buildTree(PrototypeSet data, double[] classProbs, PrototypeSet header, double minNum, boolean debug, int[] attIndicesWindow, int depth, boolean allow) Recursively generates a tree.
protected double	distribution(double[][] props, double[][][] dists, int att, PrototypeSet data) Computes class distribution for an attribute.
double[]	distributionForInstance(Prototype instance) Computes class distribution of an instance using the decision tree.
protected double	gain(double[][] dist, double priorVal) Computes value of splitting criterion after split.
boolean	getAllowUnclassifiedInstances() Get the value of NumFolds.
int	getKValue() Get the value of K.
int	getMaxDepth() Get the maximum depth of trh tree, 0 for unlimited.
double	getMinNum() Get the value of MinNum.
int	getNumFolds() Get the value of NumFolds.
int	getSeed() Gets the seed for the random number generations
java.lang.String	globalInfo() Returns a string describing classifier
java.lang.String	KValueTipText() Returns the tip text for this property
static void	main(java.lang.String[] argv) Main method for this class.
java.lang.String	maxDepthTipText() Returns the tip text for this property
java.lang.String	minNumTipText() Returns the tip text for this property
java.lang.String	numFoldsTipText() Returns the tip text for this property
int	numNodes() Computes size of the tree.
protected double	priorVal(double[][] dist) Computes value of splitting criterion before split.
java.lang.String	seedTipText() Returns the tip text for this property
void	setAllowUnclassifiedInstances(boolean newAllowUnclassifiedInstances) Set the value of AllowUnclassifiedInstances.
void	setKValue(int k) Set the value of K.
void	setMaxDepth(int value) Set the maximum depth of the tree, 0 for unlimited.
void	setMinNum(double newMinNum) Set the value of MinNum.
void	setNumFolds(int newNumFolds) Set the value of NumFolds.
void	setSeed(int seed) Set the seed for random number generation.
protected PrototypeSet[]	splitData(PrototypeSet data) Splits instances into subsets based on the given split.

Methods inherited from class java.lang.Object

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

Field Detail

m_Successors

protected RandomTree[] m_Successors

The subtrees appended to this tree.

m_Attribute

protected int m_Attribute

The attribute to split on.

m_SplitPoint

protected double m_SplitPoint

The split point.

m_Info

protected PrototypeSet m_Info

The header information.

m_Prop

protected double[] m_Prop

The proportions of training instances going down each branch.

m_ClassDistribution

protected double[] m_ClassDistribution

Class probabilities from the training data.

m_MinNum

protected double m_MinNum

Minimum number of instances for leaf.

m_KValue

protected int m_KValue

The number of attributes considered for a split.

m_randomSeed

protected int m_randomSeed

The random seed to use.

m_MaxDepth

protected int m_MaxDepth

The maximum depth of the tree (0 = unlimited)

m_NumFolds

protected int m_NumFolds

Determines how much data is used for backfitting

m_Debug

protected boolean m_Debug

Debug flag.

nClasses

protected int nClasses

Number of classes.

m_AllowUnclassifiedInstances

protected boolean m_AllowUnclassifiedInstances

Whether unclassified instances are allowed

Constructor Detail

RandomTree

public RandomTree()

Method Detail

globalInfo

public java.lang.String globalInfo()

Returns a string describing classifier

Returns:

a description suitable for displaying in the explorer/experimenter gui

minNumTipText

public java.lang.String minNumTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getMinNum

public double getMinNum()

Get the value of MinNum.

Returns:

Value of MinNum.

setMinNum

public void setMinNum(double newMinNum)

Set the value of MinNum.

Parameters:

newMinNum - Value to assign to MinNum.

KValueTipText

public java.lang.String KValueTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getKValue

public int getKValue()

Get the value of K.

Returns:

Value of K.

setKValue

public void setKValue(int k)

Set the value of K.

Parameters:

k - Value to assign to K.

seedTipText

public java.lang.String seedTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

setSeed

public void setSeed(int seed)

Set the seed for random number generation.

Parameters:

seed - the seed

getSeed

public int getSeed()

Gets the seed for the random number generations

Returns:

the seed for the random number generation

maxDepthTipText

public java.lang.String maxDepthTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getMaxDepth

public int getMaxDepth()

Get the maximum depth of trh tree, 0 for unlimited.

Returns:

the maximum depth.

numFoldsTipText

public java.lang.String numFoldsTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getNumFolds

public int getNumFolds()

Get the value of NumFolds.

Returns:

Value of NumFolds.

setNumFolds

public void setNumFolds(int newNumFolds)

Set the value of NumFolds.

Parameters:

newNumFolds - Value to assign to NumFolds.

allowUnclassifiedInstancesTipText

public java.lang.String allowUnclassifiedInstancesTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getAllowUnclassifiedInstances

public boolean getAllowUnclassifiedInstances()

Get the value of NumFolds.

Returns:

Value of NumFolds.

setAllowUnclassifiedInstances

public void setAllowUnclassifiedInstances(boolean newAllowUnclassifiedInstances)

Set the value of AllowUnclassifiedInstances.

Parameters:

newAllowUnclassifiedInstances - Value to assign to AllowUnclassifiedInstances.

setMaxDepth

public void setMaxDepth(int value)

Set the maximum depth of the tree, 0 for unlimited.

Parameters:

value - the maximum depth.

buildClassifier

public void buildClassifier(PrototypeSet data)
                     throws java.lang.Exception

Builds classifier.

Parameters:

data - the data to train with

Throws:

java.lang.Exception - if something goes wrong or the data doesn't fit

backfitData

public void backfitData(PrototypeSet data)
                 throws java.lang.Exception

Backfits the given data into the tree.

Parameters:

data - given data to backfit into the tree.

Throws:

java.lang.Exception

distributionForInstance

public double[] distributionForInstance(Prototype instance)
                                 throws java.lang.Exception

Computes class distribution of an instance using the decision tree.

Parameters:

instance - the instance to compute the distribution for

Returns:

the computed class distribution

Throws:

java.lang.Exception - if computation fails

backfitData

protected void backfitData(PrototypeSet data,
                           double[] classProbs)
                    throws java.lang.Exception

Recursively backfits data into the tree.

Parameters:

data - the data to work with

classProbs - the class distribution

Throws:

java.lang.Exception - if generation fails

buildTree

protected void buildTree(PrototypeSet data,
                         double[] classProbs,
                         PrototypeSet header,
                         double minNum,
                         boolean debug,
                         int[] attIndicesWindow,
                         int depth,
                         boolean allow)
                  throws java.lang.Exception

Recursively generates a tree.

Parameters:

data - the data to work with

classProbs - the class distribution

header - the header of the data

minNum - the minimum number of instances per leaf

debug - whether debugging is on

attIndicesWindow - the attribute window to choose attributes from

depth - the current depth

allow - if unclassified instances are allowed.

Throws:

java.lang.Exception - if generation fails

numNodes

public int numNodes()

Computes size of the tree.

Returns:

the number of nodes

splitData

protected PrototypeSet[] splitData(PrototypeSet data)
                            throws java.lang.Exception

Splits instances into subsets based on the given split.

Parameters:

data - the data to work with

Returns:

the subsets of instances

Throws:

java.lang.Exception - if something goes wrong

distribution

protected double distribution(double[][] props,
                              double[][][] dists,
                              int att,
                              PrototypeSet data)
                       throws java.lang.Exception

Computes class distribution for an attribute.

Parameters:

props -

dists -

att - the attribute index

data - the data to work with

Returns:

distribution for an attribute.

Throws:

java.lang.Exception - if something goes wrong

priorVal

protected double priorVal(double[][] dist)

Computes value of splitting criterion before split.

Parameters:

dist - the distributions

Returns:

the splitting criterion

gain

protected double gain(double[][] dist,
                      double priorVal)

Computes value of splitting criterion after split.

Parameters:

dist - the distributions

priorVal - the splitting criterion

Returns:

the gain after the split

main

public static void main(java.lang.String[] argv)

Main method for this class.

Parameters:

argv - the commandline parameters