edu.illinois.cs.cogcomp.lbjava.learn

Class BiasedRandomWeightVector

java.lang.Object

edu.illinois.cs.cogcomp.lbjava.learn.SparseWeightVector

edu.illinois.cs.cogcomp.lbjava.learn.RandomWeightVector

edu.illinois.cs.cogcomp.lbjava.learn.BiasedRandomWeightVector

All Implemented Interfaces:

Serializable, Cloneable

public class BiasedRandomWeightVector
extends RandomWeightVector

Same as the RandomWeightVector class that it extends, except that this vector also contains a bias term (also initialized randomly) which is added to every dot product and affected by every vector addition operation.

Author:

Nick Rizzolo

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected double	bias The current bias weight.

Fields inherited from class edu.illinois.cs.cogcomp.lbjava.learn.RandomWeightVector

defaultStddev, instanceNumber, random, stddev

Fields inherited from class edu.illinois.cs.cogcomp.lbjava.learn.SparseWeightVector

defaultCapacity, defaultWeight, weights

Constructor Summary

Constructors

Constructor and Description
BiasedRandomWeightVector() Instantiates this biased vector with a random bias.
BiasedRandomWeightVector(double s) Sets the specified standard deviation and a random bias.

Method Summary

Modifier and Type	Method and Description
void	clear() Empties the weight map and resets the random number generator.
double	dot(int[] exampleFeatures, double[] exampleValues, double defaultW) Takes the dot product of this BiasedRandomWeightVector with the argument vector, using the specified default weight when one is not yet present in this vector.
SparseWeightVector	emptyClone() Returns a new, empty weight vector with the same parameter settings as this one.
void	scaledAdd(int[] exampleFeatures, double[] exampleValues, double factor, double defaultW) Self-modifying vector addition where the argument vector is first scaled by the given factor.
void	write(edu.illinois.cs.cogcomp.core.datastructures.vectors.ExceptionlessOutputStream out) Writes the weight vector's internal representation in binary form.
void	write(PrintStream out) Outputs a textual representation of this vector to the specified stream.
void	write(PrintStream out, Lexicon lex) Outputs a textual representation of this vector to the specified stream.

Methods inherited from class edu.illinois.cs.cogcomp.lbjava.learn.RandomWeightVector

getWeight, read

Methods inherited from class edu.illinois.cs.cogcomp.lbjava.learn.SparseWeightVector

clone, dot, getWeight, pairwiseMultiply, readWeightVector, scaledAdd, scaledAdd, scaledMultiply, scaledMultiply, setWeight, setWeight, size, toString, toString, toStringJustWeights, toStringJustWeights

Methods inherited from class java.lang.Object

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

Field Detail

bias

protected double bias

The current bias weight.

Constructor Detail

BiasedRandomWeightVector

public BiasedRandomWeightVector()

Instantiates this biased vector with a random bias.

BiasedRandomWeightVector

public BiasedRandomWeightVector(double s)

Sets the specified standard deviation and a random bias.

Parameters:

s - The standard deviation.

Method Detail

dot

public double dot(int[] exampleFeatures,
                  double[] exampleValues,
                  double defaultW)

Takes the dot product of this BiasedRandomWeightVector with the argument vector, using the specified default weight when one is not yet present in this vector.

Overrides:

dot in class SparseWeightVector

Parameters:

exampleFeatures - The example's array of feature indices

exampleValues - The example's array of feature values

defaultW - The default weight.

Returns:

The computed dot product.

scaledAdd

public void scaledAdd(int[] exampleFeatures,
                      double[] exampleValues,
                      double factor,
                      double defaultW)

Self-modifying vector addition where the argument vector is first scaled by the given factor.

Overrides:

scaledAdd in class SparseWeightVector

Parameters:

exampleFeatures - The example's array of feature indices

exampleValues - The example's array of feature values

factor - The scaling factor.

defaultW - An initial weight for previously unseen features.

clear

public void clear()

Empties the weight map and resets the random number generator. This means that the same "random" values will be filled in for the weights if the same calls to dot(int[],double[],double) and scaledAdd(int[],double[],double,double) are made in the same order.

Overrides:

clear in class RandomWeightVector

write

public void write(PrintStream out)

Outputs a textual representation of this vector to the specified stream. The string representation is the same as in SparseWeightVector.write(PrintStream), with two added lines just after the "Begin" annotation that give the values of RandomWeightVector.stddev and bias.

Overrides:

write in class RandomWeightVector

Parameters:

out - The stream to write to.

write

public void write(PrintStream out,
                  Lexicon lex)

Outputs a textual representation of this vector to the specified stream. The string representation is the same as in SparseWeightVector.write(PrintStream), with two added lines just after the "Begin" annotation that give the values of RandomWeightVector.stddev and bias.

Overrides:

write in class RandomWeightVector

Parameters:

out - The stream to write to.

lex - The feature lexicon.

write

public void write(edu.illinois.cs.cogcomp.core.datastructures.vectors.ExceptionlessOutputStream out)

Writes the weight vector's internal representation in binary form.

Overrides:

write in class RandomWeightVector

Parameters:

out - The output stream.

emptyClone

public SparseWeightVector emptyClone()

Returns a new, empty weight vector with the same parameter settings as this one.

Overrides:

emptyClone in class RandomWeightVector

Returns:

An empty weight vector.