probabilityMachine.vpm
Class VPMBartsRMI2

java.lang.Object
  |
  +--weka.classifiers.Classifier
        |
        +--probabilityMachine.VennProbabilityClassifier
              |
              +--probabilityMachine.vpm.VPMBartsRMI2

All Implemented Interfaces:

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

public class VPMBartsRMI2

extends VennProbabilityClassifier

implements OptionHandler

The second VPM version of the BartsRMI algorithm. Uses Gaussians to model strangeness types! (Only for use on Ovarian Cancer data)

Version:

$Revision: 1.0 $

Author:

David Lindsay (davidl@cs.rhul.ac.uk)

See Also:

Serialized Form

Constructor Summary

VPMBartsRMI2()
The amazing VPMBartsRMI classifier

VPMBartsRMI2(double rmit, int vennt)
The amazing VPMBartsRMI2 classifier, used to show how to adapt a classifer to the VPM meta learning framework!

Method Summary

double[]	addDoubleToEndOfArray(double[] array, double d)
int	assignVennType(double dist, int chosenGaussian) This is the function that sets the type depending on the number of deviations from the mean! (This need not be symmetric about the mean!)
void	buildClassifier(Instances instances) Generates the classifier.
double	calculateRMI(Instance instance) Calculates the Risk of Malignancy Index find ref ? (Jacobs et al)
int[]	countNumberInSameTypeAsNewTest(int[] closestGaussianVennType) Simple function to count the number of training examples of the same Venn type as the new test example.
int[]	createSymmetricVennTranslationVector() Creates a vector to work out how to translate identifier -> index
java.lang.String	createTypeDetailString(int[] closestGaussianVennType, int[] typeDetails, int classAssigned) Output type membership details
int[]	findVennTypes(double benignAlphaMean, double benignAlphaVar, double malignantAlphaMean, double malignantAlphaVar, double[] allAlphas, double alphaNew) Computes the Venn types for training and new test example
int[]	getGaussianStdTypes(int[] closestGaussianVennType, double benignAlphaMean, double benignAlphaVar, double malignantAlphaMean, double malignantAlphaVar, double[] allAlphas, double alphaNew) Finds the number of std deviations away from the mean of each Gaussian -ve represent in Benign Gaussian +ve represent in Malignant Gaussian.
int	getNumberVennTypes() Gets the number of Venn probability types used!
java.lang.String[]	getOptions() Gets the current settings of VPM.
double	getRMIThreshold() Gets the currently set RMI threshold
java.util.Enumeration	listOptions() Returns an enumeration describing the available options
double	mahanobolisDistanceFrom(double mean, double var, double x) Finds the number of standard deviations from the mean of a Gaussian
static void	main(java.lang.String[] argv) Main method for testing this class.
java.lang.String	printArray(double[] array) Debugging function
double	probGaussian(double mean, double var, double x) Computes prob from a Gaussian!
int[]	setAsTrueGaussian(int chosenGaussianOfTest) Debugging class that sets the Gaussian to its true class
void	setNumberVennTypes(int vennt) Sets the number of Venn probability types used!
void	setOptions(java.lang.String[] options) Parses a given list of options.
void	setRMIThreshold(double rmit) Sets the RMI threshold
java.lang.String	toString() Returns a description of this classifier.
Matrix	vennProbsForInstance(Instance instance) Returns the Venn probability matrix for a given test instance.

Methods inherited from class probabilityMachine.VennProbabilityClassifier

classifyInstance, computeRowOfVPMMatrix, distributionForInstance, printArray, returnDistribution, returnUpperAndLowerProbability

Methods inherited from class weka.classifiers.Classifier

debugTipText, forName, getDebug, makeCopies, setDebug

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

VPMBartsRMI2

public VPMBartsRMI2(double rmit,
                    int vennt)

The amazing VPMBartsRMI2 classifier, used to show how to adapt a classifer to the VPM meta learning framework!

The options for this wonderful algorithm are as follows:
-r threshold
Set the threshold for the RMI (default: 250)

-y num Venn types
Sets the number of types used for the VPM

Parameters:

rmit - the RMI threshold

vennt - the number of Venn probability types used

VPMBartsRMI2

public VPMBartsRMI2()

The amazing VPMBartsRMI classifier

Method Detail

setRMIThreshold

public void setRMIThreshold(double rmit)

Sets the RMI threshold

getRMIThreshold

public double getRMIThreshold()

Gets the currently set RMI threshold

setNumberVennTypes

public void setNumberVennTypes(int vennt)

Sets the number of Venn probability types used!

getNumberVennTypes

public int getNumberVennTypes()

Gets the number of Venn probability types used!

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

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:

-r threshold
Set the threshold for the RMI (default: 250)

-y num Venn types
Sets the number of types used for the VPM

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 VPM.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class Classifier

Returns:

an array of strings suitable for passing to setOptions()

calculateRMI

public double calculateRMI(Instance instance)

Calculates the Risk of Malignancy Index find ref ? (Jacobs et al)

addDoubleToEndOfArray

public double[] addDoubleToEndOfArray(double[] array,
                                      double d)

probGaussian

public double probGaussian(double mean,
                           double var,
                           double x)

Computes prob from a Gaussian!

findVennTypes

public int[] findVennTypes(double benignAlphaMean,
                           double benignAlphaVar,
                           double malignantAlphaMean,
                           double malignantAlphaVar,
                           double[] allAlphas,
                           double alphaNew)

Computes the Venn types for training and new test example

getGaussianStdTypes

public int[] getGaussianStdTypes(int[] closestGaussianVennType,
                                 double benignAlphaMean,
                                 double benignAlphaVar,
                                 double malignantAlphaMean,
                                 double malignantAlphaVar,
                                 double[] allAlphas,
                                 double alphaNew)

Finds the number of std deviations away from the mean of each Gaussian -ve represent in Benign Gaussian +ve represent in Malignant Gaussian. Determined by the user specified number of types. eg. If 3 types specified, +3 represents in outer tails (further than 3 stdevs from mean) of Malignant Gaussian! -2 represnts in bands between 2 and 3 stdevs of mean of Benign Gaussian

assignVennType

public int assignVennType(double dist,
                          int chosenGaussian)

This is the function that sets the type depending on the number of deviations from the mean! (This need not be symmetric about the mean!)

mahanobolisDistanceFrom

public double mahanobolisDistanceFrom(double mean,
                                      double var,
                                      double x)

Finds the number of standard deviations from the mean of a Gaussian

createSymmetricVennTranslationVector

public int[] createSymmetricVennTranslationVector()

Creates a vector to work out how to translate identifier -> index

countNumberInSameTypeAsNewTest

public int[] countNumberInSameTypeAsNewTest(int[] closestGaussianVennType)

Simple function to count the number of training examples of the same Venn type as the new test example.

setAsTrueGaussian

public int[] setAsTrueGaussian(int chosenGaussianOfTest)

Debugging class that sets the Gaussian to its true class

vennProbsForInstance

public Matrix vennProbsForInstance(Instance instance)
                            throws java.lang.Exception

Returns the Venn probability matrix for a given test instance.

Specified by:

vennProbsForInstance in class VennProbabilityClassifier

Parameters:

instance - the instance to be classified

Returns:

the Venn probability matrix

Throws:

java.lang.Exception - no training instances

printArray

public java.lang.String printArray(double[] array)

Debugging function

Overrides:

printArray in class VennProbabilityClassifier

createTypeDetailString

public java.lang.String createTypeDetailString(int[] closestGaussianVennType,
                                               int[] typeDetails,
                                               int classAssigned)

Output type membership details

toString

public java.lang.String toString()

Returns a description of this classifier.

Overrides:

toString in class java.lang.Object

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 arguments for evaluation (see Evaluation).