weka.classifiers.bayes
Class ComplementNaiveBayes

java.lang.Object
  weka.classifiers.Classifier
      weka.classifiers.bayes.ComplementNaiveBayes

All Implemented Interfaces:

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

public class ComplementNaiveBayes

extends Classifier

implements OptionHandler, WeightedInstancesHandler

Class for building and using a Complement class Naive Bayes classifier. For more information see,

ICML-2003 Tackling the poor assumptions of Naive Bayes Text Classifiers P.S.: TF, IDF and length normalization transforms, as described in the paper, can be performed through weka.filters.unsupervised.StringToWordVector.

Valid options for the classifier are:

-N
Normalizes word weights for each class.

-S val
The smoothing value to use to avoid zero WordGivenClass probabilities (default 1.0).

Version:

$Revision: 1.2 $

Author:

Ashraf M. Kibriya (amk14@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

private Instances	header The instances header that'll be used in toString
private boolean	m_normalizeWordWeights True if the words weights are to be normalized
private int	numClasses Holds the number of Class values present in the set of specified instances
private double	smoothingParameter Holds the smoothing value to avoid word probabilities of zero.
private double[][]	wordWeights Weight of words for each class.

Fields inherited from class weka.classifiers.Classifier

m_Debug

Constructor Summary

ComplementNaiveBayes()

Method Summary

void	buildClassifier(Instances instances) Generates the classifier.
double	classifyInstance(Instance instance) Classifies a given instance.
boolean	getNormalizeWordWeights() Returns true if the word weights for each class are to be normalized
java.lang.String[]	getOptions() Gets the current settings of the classifier.
double	getSmoothingParameter() Gets the smoothing value to be used to avoid zero WordGivenClass probabilities.
java.lang.String	globalInfo() Returns a string describing this classifier
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	normalizeWordWeightsTipText() Returns the tip text for this property
void	setNormalizeWordWeights(boolean doNormalize) Sets whether if the word weights for each class should be normalized
void	setOptions(java.lang.String[] options) Parses a given list of options.
void	setSmoothingParameter(double val) Sets the smoothing value used to avoid zero WordGivenClass probabilities
java.lang.String	smoothingParameterTipText() Returns the tip text for this property
java.lang.String	toString() Prints out the internal model built by the classifier.

Methods inherited from class weka.classifiers.Classifier

debugTipText, distributionForInstance, forName, getDebug, makeCopies, setDebug

Methods inherited from class java.lang.Object

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

Field Detail

wordWeights

private double[][] wordWeights

Weight of words for each class. The weight is actually the log of the probability of a word (w) given a class (c) (i.e. log(Pr[w|c])). The format of the matrix is: wordWeights[class][wordAttribute]

smoothingParameter

private double smoothingParameter

Holds the smoothing value to avoid word probabilities of zero.
P.S.: According to the paper this is the Alpha i parameter

m_normalizeWordWeights

private boolean m_normalizeWordWeights

True if the words weights are to be normalized

numClasses

private int numClasses

Holds the number of Class values present in the set of specified instances

header

private Instances header

The instances header that'll be used in toString

Constructor Detail

ComplementNaiveBayes

public ComplementNaiveBayes()

Method Detail

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.

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

setOptions

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

Parses a given list of options. Valid options are:

-N
Normalizes word weights for each class.

-S val
The smoothing value to use to avoid zero WordGivenClass probabilities (default 1.0).

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

getNormalizeWordWeights

public boolean getNormalizeWordWeights()

Returns true if the word weights for each class are to be normalized

setNormalizeWordWeights

public void setNormalizeWordWeights(boolean doNormalize)

Sets whether if the word weights for each class should be normalized

normalizeWordWeightsTipText

public java.lang.String normalizeWordWeightsTipText()

Returns the tip text for this property

Returns:

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

getSmoothingParameter

public double getSmoothingParameter()

Gets the smoothing value to be used to avoid zero WordGivenClass probabilities.

setSmoothingParameter

public void setSmoothingParameter(double val)

Sets the smoothing value used to avoid zero WordGivenClass probabilities

smoothingParameterTipText

public java.lang.String smoothingParameterTipText()

Returns the tip text for this property

Returns:

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

globalInfo

public java.lang.String globalInfo()

Returns a string describing this classifier

Returns:

a description of the classifier 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 built successfully

classifyInstance

public double classifyInstance(Instance instance)
                        throws java.lang.Exception

Classifies a given instance.

The classification rule is:
MinC(forAllWords(ti*Wci))
where
ti is the frequency of word i in the given instance
Wci is the weight of word i in Class c.

For more information see section 4.4 of the paper mentioned above in the classifiers description.

Overrides:

classifyInstance in class Classifier

Parameters:

instance - the instance to be classified

Returns:

the index of the class the instance is most likely to belong.

Throws:

if - the classifier has not been built yet.

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

toString

public java.lang.String toString()

Prints out the internal model built by the classifier. In this case it prints out the word weights calculated when building the classifier.

main

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

Main method for testing this class.

Parameters:

argv - the options