LBJ2.learn
Class AdaBoost

java.lang.Object
  LBJ2.classify.Classifier
      LBJ2.learn.Learner
          LBJ2.learn.AdaBoost

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable

public class AdaBoost
extends Learner

Implementation of the AdaBoost binary classification learning algorithm. A Classifier learned with this algorithm will always produce a single Feature with value equal to a value the weak learner is capable of producing. This implementation samples from its internal distribution, giving the weak learner a new set of examples that the weak learner assumes are weighted equally.
Assumptions:

1. The Learner used as the weak learner is cloneable.
2. The weak learning algorithm is trained on objects that are each given a single label feature.

See Also:

Serialized Form

Field Summary

private double[]	alpha Parameters associated with the trained copies of the weak learner.
protected int	rounds The number of times the weak learner will be called.
protected Learner	weakLearner The weak learning algorithm to be boosted.
private Learner[]	weakLearners Will be filled with trained copies of the weak learner.

Fields inherited from class LBJ2.learn.Learner

extractor, labeler

Fields inherited from class LBJ2.classify.Classifier

containingPackage, name

Constructor Summary

AdaBoost()
This constructor is used by the LBJ2 compiler; it should never be called by a programmer.

AdaBoost(java.lang.String n, Learner w, int r)
Instantiates member variables.

Method Summary

FeatureVector	classify(java.lang.Object example) This method uses the trained parameters to make a binary decision about an example object.
java.lang.Object	clone() Returns a deep clone of this learning algorithm.
void	forget() Clears weakLearners and alpha, although this is not necessary since learn(Object[]) will overwrite them fresh each time it is called.
void	learn(java.lang.Object e) This implementation of AdaBoost cannot be trained in an online fashion, so this method simply reports an error.
void	learn(java.lang.Object[] examples) The AdaBoost training algorithm.
ScoreSet	scores(java.lang.Object example) Produces a set of scores indicating the degree to which each possible discrete classification value is associated with the given example object.
void	setExtractor(Classifier e) Sets the extractor.
void	setLabeler(Classifier l) Sets the labeler.
void	write(java.io.PrintStream out) Writes this algorithm's internal representation as text.

Methods inherited from class LBJ2.learn.Learner

doneLearning, getExtractor, getLabeler, save

Methods inherited from class LBJ2.classify.Classifier

allowableValues, binaryRead, binaryRead, binaryRead, binaryRead, binaryWrite, binaryWrite, classify, discreteValue, discreteValueArray, getCompositeChildren, getInputType, getOutputType, realValue, realValueArray, test, toString, valueIndexOf

Methods inherited from class java.lang.Object

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

Field Detail

weakLearner

protected Learner weakLearner

The weak learning algorithm to be boosted.

rounds

protected int rounds

The number of times the weak learner will be called.

weakLearners

private Learner[] weakLearners

Will be filled with trained copies of the weak learner.

alpha

private double[] alpha

Parameters associated with the trained copies of the weak learner.

Constructor Detail

AdaBoost

public AdaBoost()

This constructor is used by the LBJ2 compiler; it should never be called by a programmer.

AdaBoost

public AdaBoost(java.lang.String n,
                Learner w,
                int r)

Instantiates member variables.

Parameters:

n - The name of the classifier.

w - The weak learning algorithm.

r - The number of rounds of boosting.

Method Detail

setLabeler

public void setLabeler(Classifier l)

Sets the labeler.

Overrides:

setLabeler in class Learner

Parameters:

l - A labeling classifier.

setExtractor

public void setExtractor(Classifier e)

Sets the extractor.

Overrides:

setExtractor in class Learner

Parameters:

e - A feature extracting classifier.

learn

public void learn(java.lang.Object e)

This implementation of AdaBoost cannot be trained in an online fashion, so this method simply reports an error.

Specified by:

learn in class Learner

Parameters:

e - The parameter will be ignored.

learn

public void learn(java.lang.Object[] examples)

The AdaBoost training algorithm.

Overrides:

learn in class Learner

Parameters:

examples - The entire training set.

forget

public void forget()

Clears weakLearners and alpha, although this is not necessary since learn(Object[]) will overwrite them fresh each time it is called.

Specified by:

forget in class Learner

scores

public ScoreSet scores(java.lang.Object example)

Produces a set of scores indicating the degree to which each possible discrete classification value is associated with the given example object.

Specified by:

scores in class Learner

Parameters:

example - The object to make decisions about.

Returns:

The accumulated alpha values of weak learners that predicted the associated classification value.

classify

public FeatureVector classify(java.lang.Object example)

This method uses the trained parameters to make a binary decision about an example object.

Specified by:

classify in class Classifier

Parameters:

example - The example object.

Returns:

A binary DiscreteFeature.

write

public void write(java.io.PrintStream out)

Writes this algorithm's internal representation as text.

Specified by:

write in class Learner

Parameters:

out - The output stream.

clone

public java.lang.Object clone()

Returns a deep clone of this learning algorithm.

Overrides:

clone in class Classifier

Returns:

A shallow clone.