weka.classifiers.lazy
Class KStar

java.lang.Object
  weka.classifiers.Classifier
      weka.classifiers.lazy.KStar

All Implemented Interfaces:

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

public class KStar

extends Classifier

implements KStarConstants, UpdateableClassifier

K* is an instance-based classifier, that is the class of a test instance is based upon the class of those training instances similar to it, as determined by some similarity function. The underlying assumption of instance-based classifiers such as K*, IB1, PEBLS, etc, is that similar instances will have similar classes. For more information on K*, see

John, G. Cleary and Leonard, E. Trigg (1995) "K*: An Instance- based Learner Using an Entropic Distance Measure", Proceedings of the 12th International Conference on Machine learning, pp. 108-114.

Version:

$Revision 1.0 $

Author:

Len Trigg (len@reeltwo.com), Abdelaziz Mahoui (am14@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

protected int	m_BlendMethod 0 = use specified blend, 1 = entropic blend setting
protected KStarCache[]	m_Cache A custom data structure for caching distinct attribute values and their scale factor or stop parameter.
protected int	m_ClassType The class attribute type
protected int	m_ComputeRandomCols Flag turning on and off the computation of random class colomns
protected int	m_GlobalBlend default sphere of influence blend setting
protected int	m_InitFlag Flag turning on and off the initialisation of config variables
protected int	m_MissingMode missing value treatment
protected int	m_NumAttributes The number of attributes
protected int	m_NumClasses The number of class values
protected int	m_NumInstances The number of instances in the dataset
protected int[][]	m_RandClassCols Table of random class value colomns
protected Instances	m_Train The training instances used for classification.
static Tag[]	TAGS_MISSING Define possible missing value handling methods

Fields inherited from class weka.classifiers.Classifier

m_Debug

Fields inherited from interface weka.classifiers.lazy.kstar.KStarConstants

B_ENTROPY, B_SPHERE, EPSILON, FLOOR, FLOOR1, INITIAL_STEP, LOG2, M_AVERAGE, M_DELETE, M_MAXDIFF, M_NORMAL, NUM_RAND_COLS, OFF, ON, ROOT_FINDER_ACCURACY, ROOT_FINDER_MAX_ITER

Constructor Summary

KStar()

Method Summary

private double	attrTransProb(Instance first, Instance second, int col) Calculates the transformation probability of the indexed test attribute to the indexed train attribute.
void	buildClassifier(Instances instances) Generates the classifier.
private int[]	classValues() Note: for Nominal Class Only!
double[]	distributionForInstance(Instance instance) Calculates the class membership probabilities for the given test instance.
java.lang.String	entropicAutoBlendTipText() Returns the tip text for this property
private void	generateRandomClassColomns() Note: for Nominal Class Only!
boolean	getEntropicAutoBlend() Get whether entropic blending being used
int	getGlobalBlend() Get the value of the global blend parameter
SelectedTag	getMissingMode() Gets the method to use for handling missing values.
java.lang.String[]	getOptions() Gets the current settings of K*.
java.lang.String	globalBlendTipText() Returns the tip text for this property
java.lang.String	globalInfo() Returns a string describing classifier
private void	init_m_Attributes() Initializes the m_Attributes of the class.
private double	instanceTransformationProbability(Instance first, Instance second) Calculate the probability of the first instance transforming into the second instance: the probability is the product of the transformation probabilities of the attributes normilized over the number of instances used.
java.util.Enumeration	listOptions() Returns an enumeration describing the available options.
static void	main(java.lang.String[] argv) Main method for testing this class.
java.lang.String	missingModeTipText() Returns the tip text for this property
private int[]	randomize(int[] array, java.util.Random generator) Returns a copy of the array with its elements randomly redistributed.
void	setEntropicAutoBlend(boolean e) Set whether entropic blending is to be used.
void	setGlobalBlend(int b) Set the global blend parameter
void	setMissingMode(SelectedTag newMode) Sets the method to use for handling missing values.
void	setOptions(java.lang.String[] options) Parses a given list of options.
java.lang.String	toString() Returns a description of this classifier.
private void	update_m_Attributes() Updates the m_attributes of the class.
void	updateClassifier(Instance instance) Adds the supplied instance to the training set

Methods inherited from class weka.classifiers.Classifier

classifyInstance, debugTipText, forName, getDebug, makeCopies, setDebug

Methods inherited from class java.lang.Object

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

Field Detail

m_Train

protected Instances m_Train

The training instances used for classification.

m_NumInstances

protected int m_NumInstances

The number of instances in the dataset

m_NumClasses

protected int m_NumClasses

The number of class values

m_NumAttributes

protected int m_NumAttributes

The number of attributes

m_ClassType

protected int m_ClassType

The class attribute type

m_RandClassCols

protected int[][] m_RandClassCols

Table of random class value colomns

m_ComputeRandomCols

protected int m_ComputeRandomCols

Flag turning on and off the computation of random class colomns

m_InitFlag

protected int m_InitFlag

Flag turning on and off the initialisation of config variables

m_Cache

protected KStarCache[] m_Cache

A custom data structure for caching distinct attribute values and their scale factor or stop parameter.

m_MissingMode

protected int m_MissingMode

missing value treatment

m_BlendMethod

protected int m_BlendMethod

0 = use specified blend, 1 = entropic blend setting

m_GlobalBlend

protected int m_GlobalBlend

default sphere of influence blend setting

TAGS_MISSING

public static final Tag[] TAGS_MISSING

Define possible missing value handling methods

Constructor Detail

KStar

public KStar()

Method Detail

globalInfo

public java.lang.String globalInfo()

Returns a string describing classifier

Returns:

a description suitable for displaying in the explorer/experimenter gui

buildClassifier

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

Generates the classifier.

Specified by:

buildClassifier in class Classifier

Parameters:

instances - set of instances serving as training data

Throws:

java.lang.Exception - if the classifier has not been generated successfully

updateClassifier

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

Adds the supplied instance to the training set

Specified by:

updateClassifier in interface UpdateableClassifier

Parameters:

instance - the instance to add

Throws:

java.lang.Exception - if instance could not be incorporated successfully

distributionForInstance

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

Calculates the class membership probabilities for the given test instance.

Overrides:

distributionForInstance in class Classifier

Parameters:

instance - the instance to be classified

Returns:

predicted class probability distribution

Throws:

java.lang.Exception - if an error occurred during the prediction

instanceTransformationProbability

private double instanceTransformationProbability(Instance first,
                                                 Instance second)

Calculate the probability of the first instance transforming into the second instance: the probability is the product of the transformation probabilities of the attributes normilized over the number of instances used.

Parameters:

first - the test instance

second - the train instance

Returns:

transformation probability value

attrTransProb

private double attrTransProb(Instance first,
                             Instance second,
                             int col)

Calculates the transformation probability of the indexed test attribute to the indexed train attribute.

Parameters:

first - the test instance.

second - the train instance.

col - the index of the attribute in the instance.

Returns:

the value of the transformation probability.

missingModeTipText

public java.lang.String missingModeTipText()

Returns the tip text for this property

Returns:

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

getMissingMode

public SelectedTag getMissingMode()

Gets the method to use for handling missing values. Will be one of M_NORMAL, M_AVERAGE, M_MAXDIFF or M_DELETE.

Returns:

the method used for handling missing values.

setMissingMode

public void setMissingMode(SelectedTag newMode)

Sets the method to use for handling missing values. Values other than M_NORMAL, M_AVERAGE, M_MAXDIFF and M_DELETE will be ignored.

Parameters:

newMode - the method to use for handling missing values.

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.

globalBlendTipText

public java.lang.String globalBlendTipText()

Returns the tip text for this property

Returns:

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

setGlobalBlend

public void setGlobalBlend(int b)

Set the global blend parameter

Parameters:

b - the value for global blending

getGlobalBlend

public int getGlobalBlend()

Get the value of the global blend parameter

Returns:

the value of the global blend parameter

entropicAutoBlendTipText

public java.lang.String entropicAutoBlendTipText()

Returns the tip text for this property

Returns:

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

setEntropicAutoBlend

public void setEntropicAutoBlend(boolean e)

Set whether entropic blending is to be used.

Parameters:

e - true if entropic blending is to be used

getEntropicAutoBlend

public boolean getEntropicAutoBlend()

Get whether entropic blending being used

Returns:

true if entropic blending is used

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 K*.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class Classifier

Returns:

an array of strings suitable for passing to setOptions()

toString

public java.lang.String toString()

Returns a description of this classifier.

Returns:

a description of this classifier as a string.

main

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

Main method for testing this class.

Parameters:

argv - should contain command line options (see setOptions)

init_m_Attributes

private void init_m_Attributes()

Initializes the m_Attributes of the class.

update_m_Attributes

private void update_m_Attributes()

Updates the m_attributes of the class.

generateRandomClassColomns

private void generateRandomClassColomns()

Note: for Nominal Class Only! Generates a set of random versions of the class colomn.

classValues

private int[] classValues()

Note: for Nominal Class Only! Returns an array of the class values

Returns:

an array of class values

randomize

private int[] randomize(int[] array,
                        java.util.Random generator)

Returns a copy of the array with its elements randomly redistributed.

Parameters:

array - the array to randomize.

Returns:

a copy of the array with its elements randomly redistributed.