edu.illinois.cs.cogcomp.lbjava.learn

Class Softmax

java.lang.Object

edu.illinois.cs.cogcomp.lbjava.learn.Normalizer

edu.illinois.cs.cogcomp.lbjava.learn.Softmax

public class Softmax
extends Normalizer

The softmax normalization function replaces each score with the fraction of its exponential out of the sum of all scores' exponentials. In other words, each score si is replaced by exp(alpha si) / sumj exp(alpha sj), where alpha is a user-specified constant.

Author:

Nick Rizzolo

Field Summary

Fields

Modifier and Type	Field and Description
protected double	alpha The user-specified constant described above.

Constructor Summary

Constructors

Constructor and Description
Softmax() Default constructor; sets alpha to 1.
Softmax(double a) Initializing constructor.

Method Summary

Modifier and Type	Method and Description
double	getAlpha() Retrieves the value of alpha.
ScoreSet	normalize(ScoreSet scores) Normalizes the given ScoreSet; its scores are modified in place before it is returned.

Methods inherited from class java.lang.Object

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

Field Detail

alpha

protected double alpha

The user-specified constant described above.

Constructor Detail

Softmax

public Softmax()

Default constructor; sets alpha to 1.

Softmax

public Softmax(double a)

Initializing constructor.

Parameters:

a - The setting for alpha.

Method Detail

getAlpha

public double getAlpha()

Retrieves the value of alpha.

normalize

public ScoreSet normalize(ScoreSet scores)

Normalizes the given ScoreSet; its scores are modified in place before it is returned.

Specified by:

normalize in class Normalizer

Parameters:

scores - The set of scores to normalize.

Returns:

The normalized set of scores.