cc.mallet.fst.semi_supervised
Class GECriterion

java.lang.Object
  cc.mallet.fst.semi_supervised.GECriterion

Direct Known Subclasses:

GEKLCriterion, GEL2Criterion

public abstract class GECriterion
extends java.lang.Object

Field Summary

protected double	count
protected double[]	expectation
protected java.lang.String	name
protected double[]	target
protected double	weight

Constructor Summary

GECriterion(java.lang.String name, double[] target, double weight)

Method Summary

double	getCount() Returns the count of the feature.
double[]	getExpectation() Returns the model expectation of the feature.
double	getExpectation(int li) Returns the model expectation for the feature and label li.
protected abstract double	getGradientConstant(int labelIndex) Returns the constant value from the gradient, which will be different for different criteria.
java.lang.String	getName() Returns the constraint name.
double[]	getTarget() Returns the target expectation for the feature.
double	getTarget(int li) Returns the target expectation for the feature and label li.
double	getWeight() Returns the weight (gamma) for the constraint.
protected void	setCount(double count)
protected void	setExpectation(double[] expectation)

Methods inherited from class java.lang.Object

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

Field Detail

name

protected java.lang.String name

weight

protected double weight

target

protected double[] target

expectation

protected double[] expectation

count

protected double count

Constructor Detail

GECriterion

public GECriterion(java.lang.String name,
                   double[] target,
                   double weight)

Method Detail

getName

public java.lang.String getName()

Returns the constraint name.

getWeight

public double getWeight()

Returns the weight (gamma) for the constraint.

getTarget

public double[] getTarget()

Returns the target expectation for the feature.

getTarget

public double getTarget(int li)

Returns the target expectation for the feature and label li.

Parameters:

li - label index

getExpectation

public double[] getExpectation()

Returns the model expectation of the feature.

getExpectation

public double getExpectation(int li)

Returns the model expectation for the feature and label li.

Parameters:

li - label index

setExpectation

protected void setExpectation(double[] expectation)

getCount

public double getCount()

Returns the count of the feature.

setCount

protected void setCount(double count)

getGradientConstant

protected abstract double getGradientConstant(int labelIndex)

Returns the constant value from the gradient, which will be different for different criteria.