LBJ2.learn
Class Learner

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

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable

Direct Known Subclasses:

AdaBoost, LinearThresholdUnit, MuxLearner, NaiveBayes, SparseNetworkLearner, StochasticGradientDescent, WekaWrapper

public abstract class Learner
extends Classifier

Extend this class to create a new Classifier that learns to mimic one an oracle classifier given a feature extracting classifier and example objects.

See Also:

Serialized Form

Field Summary

protected Classifier	extractor Stores the classifiers used to produce features.
protected Classifier	labeler Stores the classifier used to produce labels.

Fields inherited from class LBJ2.classify.Classifier

containingPackage, name

Constructor Summary

protected	Learner() This constructor is used by the LBJ2 compiler; it should never be called by a programmer.
protected	Learner(java.lang.String n) Initializes the name.
protected	Learner(java.lang.String n, Classifier e) Constructor for unsupervised learning.
protected	Learner(java.lang.String n, Classifier l, Classifier e) Constructor for supervised learning.

Method Summary

void	doneLearning() Overridden by subclasses to perform any required post-processing computations after all training examples have been observed through learn(Object) and learn(Object[]).
abstract void	forget() Overridden by subclasses to reinitialize the learner to the state it started at before any learning was performed.
Classifier	getExtractor() Returns the extractor.
Classifier	getLabeler() Returns the labeler.
abstract void	learn(java.lang.Object example) Trains the learning algorithm given an object as an example.
void	learn(java.lang.Object[] examples) Trains the learning algorithm given many objects as examples.
void	save() Overridden by subclasses to store a binary representation of this learner in a pre-defined location.
abstract 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.
abstract void	write(java.io.PrintStream out) Writes the algorithm's internal representation as text.

Methods inherited from class LBJ2.classify.Classifier

allowableValues, binaryRead, binaryRead, binaryRead, binaryRead, binaryWrite, binaryWrite, classify, classify, clone, 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

labeler

protected Classifier labeler

Stores the classifier used to produce labels.

extractor

protected Classifier extractor

Stores the classifiers used to produce features.

Constructor Detail

Learner

protected Learner()

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

Learner

protected Learner(java.lang.String n)

Initializes the name.

Parameters:

n - The name of the classifier.

Learner

protected Learner(java.lang.String n,
                  Classifier e)

Constructor for unsupervised learning.

Parameters:

n - The name of the classifier.

e - The feature extracting classifier.

Learner

protected Learner(java.lang.String n,
                  Classifier l,
                  Classifier e)

Constructor for supervised learning.

Parameters:

n - The name of the classifier.

l - The labeling classifier.

e - The feature extracting classifier.

Method Detail

getLabeler

public Classifier getLabeler()

Returns the labeler.

setLabeler

public void setLabeler(Classifier l)

Sets the labeler.

Parameters:

l - A labeling classifier.

getExtractor

public Classifier getExtractor()

Returns the extractor.

setExtractor

public void setExtractor(Classifier e)

Sets the extractor.

Parameters:

e - A feature extracting classifier.

learn

public abstract void learn(java.lang.Object example)

Trains the learning algorithm given an object as an example.

Parameters:

example - An example of the desired learned classifier's behavior.

learn

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

Trains the learning algorithm given many objects as examples. This implementation simply calls learn(Object) on each of the objects in the input array. It should be overridden if there is a more efficient implementation.

Parameters:

examples - Examples of the desired learned classifier's behavior.

doneLearning

public void doneLearning()

Overridden by subclasses to perform any required post-processing computations after all training examples have been observed through learn(Object) and learn(Object[]). By default this method does nothing.

forget

public abstract void forget()

Overridden by subclasses to reinitialize the learner to the state it started at before any learning was performed.

save

public void save()

Overridden by subclasses to store a binary representation of this learner in a pre-defined location. By default this method does nothing.

scores

public abstract 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. Learners that return a real feature or more than one feature may implement this method by simply returning null.

Parameters:

example - The object to make decisions about.

Returns:

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

write

public abstract void write(java.io.PrintStream out)

Writes the algorithm's internal representation as text.

Parameters:

out - The output stream.