cc.mallet.fst.semi_supervised.pr
Class ConstraintsOptimizableByPR

java.lang.Object
  cc.mallet.fst.semi_supervised.pr.ConstraintsOptimizableByPR

All Implemented Interfaces:

Optimizable, Optimizable.ByGradientValue, java.io.Serializable

public class ConstraintsOptimizableByPR
extends java.lang.Object
implements java.io.Serializable, Optimizable.ByGradientValue

Optimizable for E-step/I-projection in Posterior Regularization (PR).

Author:

Kedar Bellare, Gregory Druck

See Also:

Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from interface cc.mallet.optimize.Optimizable

Optimizable.ByBatchGradient, Optimizable.ByCombiningBatchGradient, Optimizable.ByGISUpdate, Optimizable.ByGradient, Optimizable.ByGradientValue, Optimizable.ByHessian, Optimizable.ByValue, Optimizable.ByVotedPerceptron

Field Summary

protected double[][][][]	cachedDots
protected double[]	cachedGradient
protected double	cachedValue
protected boolean	cacheStale
protected CRF	crf
protected java.util.concurrent.ThreadPoolExecutor	executor
protected int	numParameters
protected int	numThreads
protected InstanceList	trainingSet

Constructor Summary

ConstraintsOptimizableByPR(CRF crf, InstanceList ilist, PRAuxiliaryModel model)

ConstraintsOptimizableByPR(CRF crf, InstanceList ilist, PRAuxiliaryModel model, int numThreads)

Method Summary

void	cacheDotProducts()
PRAuxiliaryModel	getAuxModel()
double[][][][]	getCachedDots()
double	getCompleteValueContribution()
protected double	getExpectationValue()
int	getNumParameters()
double	getParameter(int index)
void	getParameters(double[] params)
double	getValue() Returns the log probability of the training sequence labels and the prior over parameters.
void	getValueGradient(double[] buffer)
void	setParameter(int index, double value)
void	setParameters(double[] params)
void	shutdown()

Methods inherited from class java.lang.Object

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

Field Detail

cacheStale

protected boolean cacheStale

numParameters

protected int numParameters

numThreads

protected int numThreads

trainingSet

protected InstanceList trainingSet

cachedValue

protected double cachedValue

cachedGradient

protected double[] cachedGradient

crf

protected CRF crf

executor

protected java.util.concurrent.ThreadPoolExecutor executor

cachedDots

protected double[][][][] cachedDots

Constructor Detail

ConstraintsOptimizableByPR

public ConstraintsOptimizableByPR(CRF crf,
                                  InstanceList ilist,
                                  PRAuxiliaryModel model)

ConstraintsOptimizableByPR

public ConstraintsOptimizableByPR(CRF crf,
                                  InstanceList ilist,
                                  PRAuxiliaryModel model,
                                  int numThreads)

Method Detail

cacheDotProducts

public void cacheDotProducts()

getNumParameters

public int getNumParameters()

Specified by:

getNumParameters in interface Optimizable

getParameters

public void getParameters(double[] params)

Specified by:

getParameters in interface Optimizable

getParameter

public double getParameter(int index)

Specified by:

getParameter in interface Optimizable

setParameters

public void setParameters(double[] params)

Specified by:

setParameters in interface Optimizable

setParameter

public void setParameter(int index,
                         double value)

Specified by:

setParameter in interface Optimizable

getExpectationValue

protected double getExpectationValue()

getValue

public double getValue()

Returns the log probability of the training sequence labels and the prior over parameters.

Specified by:

getValue in interface Optimizable.ByGradientValue

getCompleteValueContribution

public double getCompleteValueContribution()

getValueGradient

public void getValueGradient(double[] buffer)

Specified by:

getValueGradient in interface Optimizable.ByGradientValue

shutdown

public void shutdown()

getCachedDots

public double[][][][] getCachedDots()

getAuxModel

public PRAuxiliaryModel getAuxModel()