GenDisMixClassifier

Overview

Package

Class

Use

Tree

Deprecated

Index

Help

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: NESTED | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

de.jstacs.classifiers.differentiableSequenceScoreBased.gendismix
Class GenDisMixClassifier

java.lang.Object
  de.jstacs.classifiers.AbstractClassifier
      de.jstacs.classifiers.AbstractScoreBasedClassifier
          de.jstacs.classifiers.differentiableSequenceScoreBased.ScoreClassifier
              de.jstacs.classifiers.differentiableSequenceScoreBased.gendismix.GenDisMixClassifier

All Implemented Interfaces:: Storable, Cloneable

Direct Known Subclasses:: MSPClassifier

public class GenDisMixClassifier
extends ScoreClassifier
extends ScoreClassifier

This class implements a classifier the optimizes the following function

$\[f(\underline{\lambda}|C,D,\underline{\alpha},\underline{\beta})$

The weights $\beta_i$ have to sum to 1. For special weights the optimization turns out to be well known

if the weights are (0,1,0), one obtains maximum likelihood,
if the weights are (0,0.5,0.5), one obtains maximum a posteriori,
if the weights are (1,0,0), one obtains maximum conditional likelihood,
if the weights are (0.5,0,0.5), one obtains maximum supervised posterior,
if the $\beta_2$ =0, one obtains the generative-discriminative trade-off,
if the $\beta_2$ =0.5, one obtains the penalized generative-discriminative trade-off.

Of course, there are also some very interesting cases with other weights.

Author:: Jens Keilwagen
See Also:: LearningPrinciple, LogGenDisMixFunction

Nested Class Summary

Nested classes/interfaces inherited from class de.jstacs.classifiers.AbstractScoreBasedClassifier
`AbstractScoreBasedClassifier.DoubleTableResult`

Field Summary
`protected double[]`	`beta` The weights that determine a point of the simplex that corresponds to the function that is optimized by this classifier.
`protected LogGenDisMixFunction`	`function` The function that is optimized in this classifier.
`protected LogPrior`	`prior` The prior that is used in this classifier.

Fields inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.ScoreClassifier
`hasBeenOptimized, NOT_TRAINED_VALUE, params, score, sostream, template`

Constructor Summary
	`GenDisMixClassifier(GenDisMixClassifierParameterSet params, LogPrior prior, double[] beta, DifferentiableStatisticalModel... score)` The main constructor.
`protected`	`GenDisMixClassifier(GenDisMixClassifierParameterSet params, LogPrior prior, double lastScore, double[] beta, DifferentiableSequenceScore... score)` This constructor creates an instance and sets the value of the last (external) optimization.
	`GenDisMixClassifier(GenDisMixClassifierParameterSet params, LogPrior prior, double lastScore, double[] beta, DifferentiableStatisticalModel... score)` This constructor creates an instance and sets the value of the last (external) optimization.
	`GenDisMixClassifier(GenDisMixClassifierParameterSet params, LogPrior prior, double genBeta, double disBeta, double priorBeta, DifferentiableStatisticalModel... score)` This convenience constructor agglomerates the `genBeta, disBeta,` and `priorBeta` into an array and calls the main constructor.
	`GenDisMixClassifier(GenDisMixClassifierParameterSet params, LogPrior prior, LearningPrinciple key, DifferentiableStatisticalModel... score)` This convenience constructor creates an array of weights for an elementary learning principle and calls the main constructor.
	`GenDisMixClassifier(StringBuffer xml)` This is the constructor for `Storable`.

Method Summary
`GenDisMixClassifier`	`clone()`
`static GenDisMixClassifier[]`	`create(GenDisMixClassifierParameterSet params, LogPrior prior, double[] weights, DifferentiableStatisticalModel[]... functions)` This method creates an array of GenDisMixClassifiers by using the cross-product of the given `DifferentiableStatisticalModel`s.
`protected void`	`extractFurtherClassifierInfosFromXML(StringBuffer xml)` Extracts further information of a classifier from an XML representation.
`protected LogGenDisMixFunction`	`getFunction(DataSet[] data, double[][] weights)` Returns the function that should be optimized.
`protected StringBuffer`	`getFurtherClassifierInfos()` This method returns further information of a classifier as a `StringBuffer`.
`String`	`getInstanceName()` Returns a short description of the classifier.
`int`	`getNumberOfThreads()` This method returns the number of used threads while optimization.
`protected String`	`getXMLTag()` Returns the `String` that is used as tag for the XML representation of the classifier.
`void`	`setNumberOfThreads(int threads)` This method allows to set the number of threads used while optimization.
`void`	`setPrior(LogPrior prior)` This method set a new prior that should be used for optimization.
`void`	`setWeights(double... beta)` This method set the weights for the summand of the function.
`String`	`toString()`

Methods inherited from class de.jstacs.classifiers.differentiableSequenceScoreBased.ScoreClassifier
`createStructure, createStructure, doOptimization, getClassifierAnnotation, getCurrentParameterSet, getDifferentiableSequenceScore, getDifferentiableSequenceScores, getLastScore, getNumericalCharacteristics, getScore, hasBeenOptimized, initUsingParameters, isInitialized, preoptimize, setOutputStream, train`

Methods inherited from class de.jstacs.classifiers.AbstractScoreBasedClassifier
`check, check, classify, classify, createDefaultClassWeights, getClassWeight, getClassWeights, getMultiClassScores, getNumberOfClasses, getPValue, getPValue, getResults, getScore, getScores, setClassWeights, setClassWeights, setThresholdClassWeights`

Methods inherited from class de.jstacs.classifiers.AbstractClassifier
`classify, evaluate, evaluate, getAlphabetContainer, getCharacteristics, getLength, toXML, train`

Methods inherited from class java.lang.Object
`equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait`

Field Detail

prior

protected LogPrior prior

The prior that is used in this classifier.

function

protected LogGenDisMixFunction function

The function that is optimized in this classifier.

beta

protected double[] beta

The weights that determine a point of the simplex that corresponds to the function that is optimized by this classifier.

Constructor Detail

GenDisMixClassifier

protected GenDisMixClassifier(GenDisMixClassifierParameterSet params,
                              LogPrior prior,
                              double lastScore,
                              double[] beta,
                              DifferentiableSequenceScore... score)
                       throws CloneNotSupportedException

This constructor creates an instance and sets the value of the last (external) optimization. This constructor should only be used if it is clear that now generative part is used in the objective function.

Parameters:: params - the common parameters; prior - the prior for the parameters of the classifier; lastScore - the score of the last optimization; beta - the weights for log likelihood, log conditional likelihood and log prior; score - the functions for the classes
Throws:: CloneNotSupportedException - if an instance could not be clone

GenDisMixClassifier

public GenDisMixClassifier(GenDisMixClassifierParameterSet params,
                           LogPrior prior,
                           double lastScore,
                           double[] beta,
                           DifferentiableStatisticalModel... score)
                    throws CloneNotSupportedException

This constructor creates an instance and sets the value of the last (external) optimization.

Parameters:: params - the common parameters; prior - the prior for the parameters of the classifier; lastScore - the score of the last optimization; beta - the weights for log likelihood, log conditional likelihood and log prior; score - the functions for the classes
Throws:: CloneNotSupportedException - if an instance could not be clone

GenDisMixClassifier

public GenDisMixClassifier(GenDisMixClassifierParameterSet params,
                           LogPrior prior,
                           double[] beta,
                           DifferentiableStatisticalModel... score)
                    throws CloneNotSupportedException

The main constructor.

Parameters:: params - the common parameters; prior - the prior for the parameters of the classifier; beta - the weights for log likelihood, log conditional likelihood and log prior; score - the functions for the classes
Throws:: CloneNotSupportedException - if an instance could not be clone

GenDisMixClassifier

public GenDisMixClassifier(GenDisMixClassifierParameterSet params,
                           LogPrior prior,
                           double genBeta,
                           double disBeta,
                           double priorBeta,
                           DifferentiableStatisticalModel... score)
                    throws CloneNotSupportedException

This convenience constructor agglomerates the genBeta, disBeta, and priorBeta into an array and calls the main constructor.

Parameters:: params - the common parameters; prior - the prior for the parameters of the classifier; genBeta - the weight for the log likelihood; disBeta - the weight for the log conditional likelihood; priorBeta - the weight for the log prior; score - the functions for the classes
Throws:: CloneNotSupportedException - if an instance could not be clone

GenDisMixClassifier

public GenDisMixClassifier(GenDisMixClassifierParameterSet params,
                           LogPrior prior,
                           LearningPrinciple key,
                           DifferentiableStatisticalModel... score)
                    throws CloneNotSupportedException

This convenience constructor creates an array of weights for an elementary learning principle and calls the main constructor.

Parameters:: params - the common parameters; prior - the prior for the parameters of the classifier; key - the key for an elementary LearningPrinciple; score - the functions for the classes
Throws:: CloneNotSupportedException - if an instance could not be clone
See Also:: LearningPrinciple.getBeta(LearningPrinciple)

GenDisMixClassifier

public GenDisMixClassifier(StringBuffer xml)
                    throws NonParsableException

This is the constructor for Storable.

Parameters:: xml - the xml representation
Throws:: NonParsableException - if the representation could not be parsed.

Method Detail

clone

public GenDisMixClassifier clone()
                          throws CloneNotSupportedException

Overrides:: clone in class ScoreClassifier

Throws:: CloneNotSupportedException

getFunction

protected LogGenDisMixFunction getFunction(DataSet[] data,
                                           double[][] weights)
                                    throws Exception

Description copied from class: ScoreClassifier

Returns the function that should be optimized.

Specified by:: getFunction in class ScoreClassifier

Parameters:: data - the data sets; weights - the weights of the sequences of the data sets
Returns:: the function that should be optimized
Throws:: Exception - if something went wrong

setPrior

public void setPrior(LogPrior prior)

This method set a new prior that should be used for optimization. Since it could not be ensured that the classifier is optimal now isTrained will return false after invoking this method.

Parameters:: prior - the new prior

setWeights

public void setWeights(double... beta)
                throws IllegalArgumentException

This method set the weights for the summand of the function. Since it could not be ensured that the classifier is optimal now isTrained will return false after invoking this method.

Parameters:: beta - the weights
Throws:: IllegalArgumentException - if the weights array is not correct
See Also:: LearningPrinciple.getBeta(LearningPrinciple)

getXMLTag

protected String getXMLTag()

Description copied from class: AbstractClassifier

Returns the String that is used as tag for the XML representation of the classifier. This method is used by the methods AbstractClassifier.fromXML(StringBuffer) and AbstractClassifier.toXML().

Specified by:: getXMLTag in class ScoreClassifier

Returns:: the String that is used as tag for the XML representation of the classifier

getFurtherClassifierInfos

protected StringBuffer getFurtherClassifierInfos()

Description copied from class: AbstractClassifier

This method returns further information of a classifier as a StringBuffer. This method is used by the method AbstractClassifier.toXML() and should not be made public.

Overrides:: getFurtherClassifierInfos in class ScoreClassifier

Returns:: further information of a classifier as a StringBuffer
See Also:: AbstractClassifier.toXML()

extractFurtherClassifierInfosFromXML

protected void extractFurtherClassifierInfosFromXML(StringBuffer xml)
                                             throws NonParsableException

Description copied from class: AbstractClassifier

Extracts further information of a classifier from an XML representation. This method is used by the method AbstractClassifier.fromXML(StringBuffer) and should not be made public.

Overrides:: extractFurtherClassifierInfosFromXML in class ScoreClassifier

Parameters:: xml - the XML representation as StringBuffer
Throws:: NonParsableException - if the information could not be parsed out of the XML representation (the StringBuffer could not be parsed)
See Also:: AbstractClassifier.fromXML(StringBuffer)

create

public static GenDisMixClassifier[] create(GenDisMixClassifierParameterSet params,
                                           LogPrior prior,
                                           double[] weights,
                                           DifferentiableStatisticalModel[]... functions)
                                    throws CloneNotSupportedException

This method creates an array of GenDisMixClassifiers by using the cross-product of the given DifferentiableStatisticalModels.

Parameters:

params - the parameters that will be used in all classifiers

prior - the prior that will be used in all classifiers

weights - the weights that will be used in all classifiers

functions - the DifferentiableStatisticalModels

functions[i] are the DifferentiableStatisticalModels that can be used for class i
functions.length has to be at least 2

Returns:

an array of GenDisMixClassifiers

Throws:

CloneNotSupportedException - if the some item could not be cloned

See Also:

GenDisMixClassifier(GenDisMixClassifierParameterSet, LogPrior, double[],
      DifferentiableStatisticalModel[])

getInstanceName

public String getInstanceName()

Description copied from class: AbstractClassifier

Returns a short description of the classifier.

Overrides:: getInstanceName in class ScoreClassifier

Returns:: a short description of the classifier

getNumberOfThreads

public int getNumberOfThreads()

This method returns the number of used threads while optimization.

Returns:: the number of used threads while optimization

toString

public String toString()

Overrides:: toString in class Object

setNumberOfThreads

public void setNumberOfThreads(int threads)
                        throws SimpleParameter.IllegalValueException

This method allows to set the number of threads used while optimization.

Parameters:: threads - the number of threads
Throws:: SimpleParameter.IllegalValueException - if the number of threads is not possible

Overview

Package

Class

Use

Tree

Deprecated

Index

Help

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: NESTED | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

de.jstacs.classifiers.differentiableSequenceScoreBased.gendismix Class GenDisMixClassifier

prior

function

beta

GenDisMixClassifier

GenDisMixClassifier

GenDisMixClassifier

GenDisMixClassifier

GenDisMixClassifier

GenDisMixClassifier

clone

getFunction

setPrior

setWeights

getXMLTag

getFurtherClassifierInfos

extractFurtherClassifierInfosFromXML

create

getInstanceName

getNumberOfThreads

toString

setNumberOfThreads

de.jstacs.classifiers.differentiableSequenceScoreBased.gendismix
Class GenDisMixClassifier