csli.util.classify.berkeley
Interface Learner

All Known Implementing Classes:

C4_5Learner, SvmLearner

public interface Learner

Author:

mpurver

Method Summary

ClassificationSet[]	classify(TrainingSet testData) Classify some test data.
void	init(String fileStem) Initialise model & output files, training and classification commands.
void	setModel(File modelFile) Set a classification model from file (as an alternative to training - although some classifiers (e.g.
String	toString(ClassificationSet data) String conversion for particular learner file format.
String	toString(FeatureSet data) String conversion for particular learner file format.
String	toString(TrainingSet data) String conversion for particular learner file format.
String	toString(TrainingType data) String conversion for particular learner file format.
boolean	train(TrainingSet trainingData) Learn a model from training data.

Method Detail

init

void init(String fileStem)

Initialise model & output files, training and classification commands.

Parameters:

fileStem - the filestem to user for model/result files

train

boolean train(TrainingSet trainingData)

Learn a model from training data. Should return false on error.

Parameters:

trainingData - the training data to learn from

setModel

void setModel(File modelFile)

Set a classification model from file (as an alternative to training - although some classifiers (e.g. Svm, C4_5) work from model files anyway, in which case initial setup does all you need). Should throw RuntimeException on error.

Parameters:

modelFile - the model file (format learner-dependent)

classify

ClassificationSet[] classify(TrainingSet testData)

Classify some test data. Should return NULL on error.

Parameters:

testData - the test data to classify

Returns:

the classification set

toString

String toString(TrainingSet data)

String conversion for particular learner file format.

Parameters:

data - the data to convert

Returns:

the String representation

toString

String toString(TrainingType data)

String conversion for particular learner file format.

Parameters:

data - the data to convert

Returns:

the String representation

toString

String toString(FeatureSet data)

String conversion for particular learner file format.

Parameters:

data - the data to convert

Returns:

the String representation

toString

String toString(ClassificationSet data)

String conversion for particular learner file format.

Parameters:

data - the data to convert

Returns:

the String representation