weka.classifiers.meta
Class RacedIncrementalLogitBoost

java.lang.Object
  weka.classifiers.Classifier
      weka.classifiers.meta.RacedIncrementalLogitBoost

All Implemented Interfaces:

java.lang.Cloneable, OptionHandler, java.io.Serializable, UpdateableClassifier

public class RacedIncrementalLogitBoost

extends Classifier

implements OptionHandler, UpdateableClassifier

Classifier for incremental learning of large datasets by way of racing logit-boosted committees. Valid options are:

-C num
Set the minimum chunk size (default 500).

-M num
Set the maximum chunk size (default 8000).

-V num
Set the validation set size (default 5000).

-D
Turn on debugging output.

-W classname
Specify the full class name of a weak learner as the basis for boosting (required).

-Q
Use resampling instead of reweighting.

-S seed
Random number seed for resampling (default 1).

-P type
The type of pruning to use.

Options after -- are passed to the designated learner.

Version:

$Revision: 1.3 $

Author:

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

See Also:

Serialized Form

Nested Class Summary

protected class

RacedIncrementalLogitBoost.Committee

Field Summary

protected RacedIncrementalLogitBoost.Committee	m_bestCommittee The current best committee
protected Attribute	m_ClassAttribute The actual class attribute (for getting class names)
protected Classifier	m_Classifier The model base classifier to use
protected FastVector	m_committees The committees
protected Instances	m_currentSet The instances currently in memory for training
protected boolean	m_Debug Whether to output debug messages
protected int	m_maxBatchSizeRequired The maximum number of instances required for processing
protected int	m_maxChunkSize The maimum chunk size used for training
protected int	m_minChunkSize The minimum chunk size used for training
protected int	m_NumClasses The number of classes
protected Instances	m_NumericClassData Dummy dataset with a numeric class
protected int	m_numInstancesConsumed The number of instances consumed
protected int	m_PruningType The pruning type used
protected java.util.Random	m_RandomInstance The random number generator used
protected int	m_Seed Seed for boosting with resampling.
protected boolean	m_UseResampling Whether to use resampling
protected int	m_validationChunkSize The size of the validation set
protected Instances	m_validationSet The instances used for validation
protected boolean	m_validationSetChanged Whether the validation set has recently been changed
protected ZeroR	m_zeroR The default scheme used when committees aren't ready
static int	PRUNETYPE_LOGLIKELIHOOD
static int	PRUNETYPE_NONE The pruning types
static Tag[]	TAGS_PRUNETYPE
protected static double	Z_MAX A threshold for responses (Friedman suggests between 2 and 4)

Constructor Summary

RacedIncrementalLogitBoost()

Method Summary

void	buildClassifier(Instances data) Builds the classifier.
java.lang.String	classifierTipText()
java.lang.String	debugTipText() Returns the tip text for this property
double[]	distributionForInstance(Instance instance) Computes class distribution of an instance using the best committee.
int	getBestCommitteeChunkSize() Get the best committee chunk size
double	getBestCommitteeErrorEstimate() Get the best committee's error on the validation data
double	getBestCommitteeLLEstimate() Get the best committee's log likelihood on the validation data
int	getBestCommitteeSize() Get the number of members in the best committee
Classifier	getClassifier() Get the classifier used as the classifier
boolean	getDebug() Get whether debugging is turned on
int	getMaxChunkSize() Get the maximum chunk size
int	getMinChunkSize() Get the minimum chunk size
java.lang.String[]	getOptions() Gets the current settings of the Classifier.
SelectedTag	getPruningType() Get the pruning type
int	getSeed() Get seed for resampling.
boolean	getUseResampling() Get whether resampling is turned on
int	getValidationChunkSize() Get the validation chunk size
java.lang.String	globalInfo()
java.util.Enumeration	listOptions() Returns an enumeration describing the available options
static void	main(java.lang.String[] argv) Main method for this class.
java.lang.String	maxChunkSizeTipText()
java.lang.String	minChunkSizeTipText()
java.lang.String	pruningTypeTipText()
protected static double	RtoP(double[] Fs, int j) Convert from function responses to probabilities
java.lang.String	seedTipText()
void	setClassifier(Classifier newClassifier) Set the classifier for boosting.
void	setDebug(boolean debug) Set debugging mode
void	setMaxChunkSize(int chunkSize) Set the maximum chunk size
void	setMinChunkSize(int chunkSize) Set the minimum chunk size
void	setOptions(java.lang.String[] options) Parses a given list of options.
void	setPruningType(SelectedTag pruneType) Set the pruning type
void	setSeed(int seed) Set seed for resampling.
void	setUseResampling(boolean r) Set resampling mode
void	setValidationChunkSize(int chunkSize) Set the validation chunk size
java.lang.String	toString() Returns description of the boosted classifier.
void	updateClassifier(Instance instance) Updates the classifier.
java.lang.String	useResamplingTipText()
java.lang.String	validationChunkSizeTipText()

Methods inherited from class weka.classifiers.Classifier

classifyInstance, forName, makeCopies

Methods inherited from class java.lang.Object

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

Field Detail

PRUNETYPE_NONE

public static final int PRUNETYPE_NONE

The pruning types

See Also:

Constant Field Values

PRUNETYPE_LOGLIKELIHOOD

public static final int PRUNETYPE_LOGLIKELIHOOD

See Also:

Constant Field Values

TAGS_PRUNETYPE

public static final Tag[] TAGS_PRUNETYPE

m_Classifier

protected Classifier m_Classifier

The model base classifier to use

m_committees

protected FastVector m_committees

The committees

m_PruningType

protected int m_PruningType

The pruning type used

m_UseResampling

protected boolean m_UseResampling

Whether to use resampling

m_Seed

protected int m_Seed

Seed for boosting with resampling.

m_NumClasses

protected int m_NumClasses

The number of classes

Z_MAX

protected static final double Z_MAX

A threshold for responses (Friedman suggests between 2 and 4)

See Also:

Constant Field Values

m_NumericClassData

protected Instances m_NumericClassData

Dummy dataset with a numeric class

m_ClassAttribute

protected Attribute m_ClassAttribute

The actual class attribute (for getting class names)

m_minChunkSize

protected int m_minChunkSize

The minimum chunk size used for training

m_maxChunkSize

protected int m_maxChunkSize

The maimum chunk size used for training

m_validationChunkSize

protected int m_validationChunkSize

The size of the validation set

m_numInstancesConsumed

protected int m_numInstancesConsumed

The number of instances consumed

m_validationSet

protected Instances m_validationSet

The instances used for validation

m_currentSet

protected Instances m_currentSet

The instances currently in memory for training

m_bestCommittee

protected RacedIncrementalLogitBoost.Committee m_bestCommittee

The current best committee

m_zeroR

protected ZeroR m_zeroR

The default scheme used when committees aren't ready

m_validationSetChanged

protected boolean m_validationSetChanged

Whether the validation set has recently been changed

m_maxBatchSizeRequired

protected int m_maxBatchSizeRequired

The maximum number of instances required for processing

m_Debug

protected boolean m_Debug

Whether to output debug messages

m_RandomInstance

protected java.util.Random m_RandomInstance

The random number generator used

Constructor Detail

RacedIncrementalLogitBoost

public RacedIncrementalLogitBoost()

Method Detail

buildClassifier

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

Builds the classifier.

Specified by:

buildClassifier in class Classifier

Parameters:

data - set of instances serving as training data

Throws:

java.lang.Exception - if something goes wrong

updateClassifier

public void updateClassifier(Instance instance)
                      throws java.lang.Exception

Updates the classifier.

Specified by:

updateClassifier in interface UpdateableClassifier

Parameters:

instance - the next instance in the stream of training data

Throws:

java.lang.Exception - if something goes wrong

RtoP

protected static double RtoP(double[] Fs,
                             int j)
                      throws java.lang.Exception

Convert from function responses to probabilities

Parameters:

j - the class value of interest

Returns:

the probability prediction for j

Throws:

java.lang.Exception

distributionForInstance

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

Computes class distribution of an instance using the best committee.

Overrides:

distributionForInstance in class Classifier

Parameters:

instance - the instance to be classified

Returns:

an array containing the estimated membership probabilities of the test instance in each class or the numeric prediction

Throws:

java.lang.Exception - if distribution could not be computed successfully

listOptions

public java.util.Enumeration listOptions()

Returns an enumeration describing the available options

Specified by:

listOptions in interface OptionHandler

Overrides:

listOptions in class Classifier

Returns:

an enumeration of all the available options

setOptions

public void setOptions(java.lang.String[] options)
                throws java.lang.Exception

Parses a given list of options. Valid options are:

Specified by:

setOptions in interface OptionHandler

Overrides:

setOptions in class Classifier

Parameters:

options - the list of options as an array of strings

Throws:

java.lang.Exception - if an option is not supported

getOptions

public java.lang.String[] getOptions()

Gets the current settings of the Classifier.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class Classifier

Returns:

an array of strings suitable for passing to setOptions

globalInfo

public java.lang.String globalInfo()

Returns:

a description of the classifier suitable for displaying in the explorer/experimenter gui

classifierTipText

public java.lang.String classifierTipText()

Returns:

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

setClassifier

public void setClassifier(Classifier newClassifier)

Set the classifier for boosting. The learner should be able to handle numeric class attributes.

Parameters:

newClassifier - the Classifier to use.

getClassifier

public Classifier getClassifier()

Get the classifier used as the classifier

Returns:

the classifier used as the classifier

minChunkSizeTipText

public java.lang.String minChunkSizeTipText()

Returns:

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

setMinChunkSize

public void setMinChunkSize(int chunkSize)

Set the minimum chunk size

Parameters:

chunkSize -

getMinChunkSize

public int getMinChunkSize()

Get the minimum chunk size

Returns:

the chunk size

maxChunkSizeTipText

public java.lang.String maxChunkSizeTipText()

Returns:

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

setMaxChunkSize

public void setMaxChunkSize(int chunkSize)

Set the maximum chunk size

Parameters:

chunkSize -

getMaxChunkSize

public int getMaxChunkSize()

Get the maximum chunk size

Returns:

the chunk size

validationChunkSizeTipText

public java.lang.String validationChunkSizeTipText()

Returns:

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

setValidationChunkSize

public void setValidationChunkSize(int chunkSize)

Set the validation chunk size

Parameters:

chunkSize -

getValidationChunkSize

public int getValidationChunkSize()

Get the validation chunk size

Returns:

the chunk size

pruningTypeTipText

public java.lang.String pruningTypeTipText()

Returns:

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

setPruningType

public void setPruningType(SelectedTag pruneType)

Set the pruning type

Parameters:

pruneType -

getPruningType

public SelectedTag getPruningType()

Get the pruning type

Returns:

the type

debugTipText

public java.lang.String debugTipText()

Description copied from class: Classifier

Returns the tip text for this property

Overrides:

debugTipText in class Classifier

Returns:

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

setDebug

public void setDebug(boolean debug)

Set debugging mode

Overrides:

setDebug in class Classifier

Parameters:

debug - true if debug output should be printed

getDebug

public boolean getDebug()

Get whether debugging is turned on

Overrides:

getDebug in class Classifier

Returns:

true if debugging output is on

useResamplingTipText

public java.lang.String useResamplingTipText()

Returns:

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

setUseResampling

public void setUseResampling(boolean r)

Set resampling mode

getUseResampling

public boolean getUseResampling()

Get whether resampling is turned on

Returns:

true if resampling output is on

seedTipText

public java.lang.String seedTipText()

Returns:

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

setSeed

public void setSeed(int seed)

Set seed for resampling.

Parameters:

seed - the seed for resampling

getSeed

public int getSeed()

Get seed for resampling.

Returns:

the seed for resampling

getBestCommitteeChunkSize

public int getBestCommitteeChunkSize()

Get the best committee chunk size

getBestCommitteeSize

public int getBestCommitteeSize()

Get the number of members in the best committee

getBestCommitteeErrorEstimate

public double getBestCommitteeErrorEstimate()

Get the best committee's error on the validation data

getBestCommitteeLLEstimate

public double getBestCommitteeLLEstimate()

Get the best committee's log likelihood on the validation data

toString

public java.lang.String toString()

Returns description of the boosted classifier.

Returns:

description of the boosted classifier as a string

main

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

Main method for this class.