Package de.fu_berlin.ties.classify

This package provides functionality for classification of texts and feature vectors.

See:
          Description

Interface Summary
Classifier Classes implementing this interface must be able to classify items represented by feature vectors.
 

Class Summary
ClassTrain Classifies a list of files, training the classifier on each error.
ExternalClassifier A proxy that provides a trainable classifier by communicating with an external (non-Java) program.
MetaClassifier A meta classifier combines several layers of classifiers.
MultiBinaryClassifier This classifier converts an multi-class classification task into a several binary (two-class) classification task.
OneAgainstTheRestClassifier This classifier converts an multi-class classification task into a several binary (two-class) classification task.
Prediction A prediction, wrapping the predicted class and the probability of the prediction.
PredictionComparator A comparison function that compares Predictions based on their probabilities.
PredictionDistribution A distribution over the classes predicted by a classifier.
Probability Wraps a probability.
Reranker Reranks the predictions in a distribution by multiplying the probabilities of each of them with a bias, if specified for the type of the prediction.
TrainableClassifier Classifiers extending this abstract class must provide a training mechanism by implementing the TrainableClassifier.doTrain(FeatureVector, String, ContextMap) method.
 

Package de.fu_berlin.ties.classify Description

This package provides functionality for classification of texts and feature vectors.



Copyright © 2003-2004 Christian Siefkes. All Rights Reserved.