jbnc.measures
Class QualityMeasure

java.lang.Object
  jbnc.measures.QualityMeasure

Direct Known Subclasses:

QualityMeasureHGC, QualityMeasureLC, QualityMeasureLCV, QualityMeasureLCV_LOO, QualityMeasureSB

public abstract class QualityMeasure

extends java.lang.Object

Implements an abstract quality measure used by search algorithms to rank Bayesian networks.

Since:

June 1, 1999

Author:

Jarek Sacha

Field Summary

protected double	alphaK Value of the prior.
protected DatasetInt	dataset
protected boolean	usePriors Flag indicating use of priors.

Constructor Summary

QualityMeasure()
Default constructor.

QualityMeasure(DatasetInt dataset)
Create a quality measure based on a dataset.

Method Summary

abstract double	evaluate(BayesianNetworks.BayesNet net) Calculate the quality measure for the given network and data set.
double	getAlphaK() Gets value of the prior.
abstract java.lang.String	getName() Gets name of the quality measure.
boolean	getUsePriors() Gets usePriors flag.
void	setAlphaK(double alphaK) Sets value of the prior.
void	setDataset(DatasetInt dataset) Sets dataset used for grading Bayesian networks.
void	setUsePriors(boolean usePriors) Sets usePriors flag.

Methods inherited from class java.lang.Object

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

Field Detail

dataset

protected DatasetInt dataset

usePriors

protected boolean usePriors

Flag indicating use of priors.

alphaK

protected double alphaK

Value of the prior.

Constructor Detail

QualityMeasure

public QualityMeasure()

Default constructor.

QualityMeasure

public QualityMeasure(DatasetInt dataset)

Create a quality measure based on a dataset.

Parameters:

dataset - Description of Parameter

Method Detail

setDataset

public void setDataset(DatasetInt dataset)

Sets dataset used for grading Bayesian networks.

Parameters:

dataset - The new Dataset value

setUsePriors

public void setUsePriors(boolean usePriors)

Sets usePriors flag.

Parameters:

usePriors - The new UsePriors value

setAlphaK

public void setAlphaK(double alphaK)

Sets value of the prior.

Parameters:

alphaK - The new AlphaK value

getName

public abstract java.lang.String getName()

Gets name of the quality measure.

Returns:

The Name value

getUsePriors

public boolean getUsePriors()

Gets usePriors flag.

Returns:

The UsePriors value

getAlphaK

public double getAlphaK()

Gets value of the prior.

Returns:

The AlphaK value

evaluate

public abstract double evaluate(BayesianNetworks.BayesNet net)
                         throws java.lang.Exception

Calculate the quality measure for the given network and data set.

Parameters:

net - Network to be evaluated.

Returns:

value of the quality measure.

Throws:

java.lang.Exception - Description of Exception