weka.classifiers.rules
Class Ridor.RidorRule

java.lang.Object
  weka.classifiers.rules.Ridor.RidorRule

All Implemented Interfaces:

WeightedInstancesHandler

Enclosing class:

Ridor

private class Ridor.RidorRule

extends java.lang.Object

implements WeightedInstancesHandler

This class implements a single rule that predicts the 2-class distribution. A rule consists of antecedents "AND"ed together and the consequent (class value) for the classification. In this case, the consequent is the distribution of the available classes (always 2 classes) in the dataset. In this class, the Information Gain (p*[log(p/t) - log(P/T)]) is used to select an antecedent and Reduced Error Prunning (REP) is used to prune the rule.

Field Summary

private double	m_AccuG The accurate and covered data of this rule in the growing data
protected FastVector	m_Antds The vector of antecedents of this rule
private double	m_Class The internal representation of the class label to be predicted
private Attribute	m_ClassAttribute The class attribute of the data
private double	m_CoverG The accurate and covered data of this rule in the growing data
private double	m_CoverP The sum of weights of the data covered by this rule in the pruning data
private double	m_Worth The worth value of this rule, in this case, accurate # in pruning data
private double	m_WorthRate The worth rate of this rule, in this case, accuracy rate in the pruning data

Constructor Summary

private

Ridor.RidorRule()

Method Summary

void	buildClassifier(Instances instances) Builds a single rule learner with REP dealing with 2 classes.
private double	computeDefAccu(Instances data) Private function to compute default number of accurate instances in the specified data for m_Class
private Instances	computeInfoGain(Instances instances, double defAcRt, Ridor.Antd antd) Compute the best information gain for the specified antecedent
Instances[]	coveredByRule(Instances insts) Find all the instances in the dataset covered by this rule.
double	getAccuG()
double	getCoverG()
double	getCoverP()
double	getPredictedClass()
double	getWorth()
double	getWorthRate() The following are get functions after prune() has set the value of worthRate and worth
private void	grow(Instances data) Build one rule using the growing data
boolean	hasAntds() Whether this rule has antecedents, i.e. whether it is a default rule
boolean	isCover(Instance datum) Whether the instance covered by this rule
private void	prune(Instances pruneData) Prune the rule using the pruning data and update the worth parameters for this rule The accuracy rate is used to prune the rule.
void	setPredictedClass(double cl) The access functions for parameters
java.lang.String	toString() Prints this rule
java.lang.String	toString(java.lang.String att, java.lang.String cl) Prints this rule with the specified class label

Methods inherited from class java.lang.Object

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

Field Detail

m_Class

private double m_Class

The internal representation of the class label to be predicted

m_ClassAttribute

private Attribute m_ClassAttribute

The class attribute of the data

m_Antds

protected FastVector m_Antds

The vector of antecedents of this rule

m_WorthRate

private double m_WorthRate

The worth rate of this rule, in this case, accuracy rate in the pruning data

m_Worth

private double m_Worth

The worth value of this rule, in this case, accurate # in pruning data

m_CoverP

private double m_CoverP

The sum of weights of the data covered by this rule in the pruning data

m_CoverG

private double m_CoverG

The accurate and covered data of this rule in the growing data

m_AccuG

private double m_AccuG

The accurate and covered data of this rule in the growing data

Constructor Detail

Ridor.RidorRule

private Ridor.RidorRule()

Method Detail

setPredictedClass

public void setPredictedClass(double cl)

The access functions for parameters

getPredictedClass

public double getPredictedClass()

buildClassifier

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

Builds a single rule learner with REP dealing with 2 classes. This rule learner always tries to predict the class with label m_Class.

Parameters:

instances - the training data

Throws:

java.lang.Exception - if classifier can't be built successfully

coveredByRule

public Instances[] coveredByRule(Instances insts)

Find all the instances in the dataset covered by this rule. The instances not covered will also be deducted from the the original data and returned by this procedure.

Parameters:

insts - the dataset to be covered by this rule.

Returns:

the instances covered and not covered by this rule

isCover

public boolean isCover(Instance datum)

Whether the instance covered by this rule

Returns:

the boolean value indicating whether the instance is covered by this rule

hasAntds

public boolean hasAntds()

Whether this rule has antecedents, i.e. whether it is a default rule

Returns:

the boolean value indicating whether the rule has antecedents

grow

private void grow(Instances data)

Build one rule using the growing data

Parameters:

data - the growing data used to build the rule

computeInfoGain

private Instances computeInfoGain(Instances instances,
                                  double defAcRt,
                                  Ridor.Antd antd)

Compute the best information gain for the specified antecedent

Parameters:

defAcRt - the default accuracy rate of data

antd - the specific antecedent

Returns:

the data covered by the antecedent

prune

private void prune(Instances pruneData)

Prune the rule using the pruning data and update the worth parameters for this rule The accuracy rate is used to prune the rule.

Parameters:

pruneData - the pruning data used to prune the rule

computeDefAccu

private double computeDefAccu(Instances data)

Private function to compute default number of accurate instances in the specified data for m_Class

Parameters:

data - the data in question

Returns:

the default accuracy number

getWorthRate

public double getWorthRate()

The following are get functions after prune() has set the value of worthRate and worth

getWorth

public double getWorth()

getCoverP

public double getCoverP()

getCoverG

public double getCoverG()

getAccuG

public double getAccuG()

toString

public java.lang.String toString(java.lang.String att,
                                 java.lang.String cl)

Prints this rule with the specified class label

Parameters:

att - the string standing for attribute in the consequent of this rule

cl - the string standing for value in the consequent of this rule

Returns:

a textual description of this rule with the specified class label

toString

public java.lang.String toString()

Prints this rule

Returns:

a textual description of this rule