Uses of Class
de.fu_berlin.ties.classify.TrainableClassifier

Packages that use TrainableClassifier
de.fu_berlin.ties.classify This package provides functionality for classification of texts and feature vectors. 
de.fu_berlin.ties.classify.winnow This package contains the Winnow classification algorithm and related algorithms and classes. 
de.fu_berlin.ties.extract This package handles information extraction and entitiy recognition. 
 

Uses of TrainableClassifier in de.fu_berlin.ties.classify
 

Subclasses of TrainableClassifier in de.fu_berlin.ties.classify
 class ExternalClassifier
          A proxy that provides a trainable classifier by communicating with an external (non-Java) program.
(package private)  class MetaClassifier
          A meta classifier combines several layers of classifiers.
 class MultiBinaryClassifier
          This classifier converts an multi-class classification task into a several binary (two-class) classification task.
 class OneAgainstTheRestClassifier
          This classifier converts an multi-class classification task into a several binary (two-class) classification task.
 

Methods in de.fu_berlin.ties.classify that return TrainableClassifier
static TrainableClassifier TrainableClassifier.createClassifier(Set<String> allValidClasses)
          Factory method that delegates to createClassifier(Set, TiesConfiguration) using the standard configuration.
static TrainableClassifier TrainableClassifier.createClassifier(Set<String> allValidClasses, File runDirectory, FeatureTransformer trans, String[] spec, TiesConfiguration conf)
          Factory method that creates a trainable classifier based on the provided specification.
static TrainableClassifier TrainableClassifier.createClassifier(Set<String> allValidClasses, File runDirectory, TiesConfiguration conf)
          Factory method that delegates to createClassifier(Set, File, FeatureTransformer, String[], TiesConfiguration).
static TrainableClassifier TrainableClassifier.createClassifier(Set<String> allValidClasses, TiesConfiguration conf)
          Factory method that delegates to createClassifier(Set, File, TiesConfiguration) without specifying an run directory.
 

Uses of TrainableClassifier in de.fu_berlin.ties.classify.winnow
 

Subclasses of TrainableClassifier in de.fu_berlin.ties.classify.winnow
 class UltraconservativeWinnow
          A combination of Winnow with the "ultraconservative" approach proposed by Koby Crammer and Yoram Singer.
 class Winnow
          Classifier implementing the Winnow algorithm (Nick Littlestone).
 

Uses of TrainableClassifier in de.fu_berlin.ties.extract
 

Constructors in de.fu_berlin.ties.extract with parameters of type TrainableClassifier
Trainer(String outExt, TargetStructure targetStruct, TrainableClassifier[] theClassifiers, Representation theRepresentation, CombinationStrategy combiStrat, TokenizerFactory tFactory, TrainableFilter sentFilter)
          Creates a new instance, using the standard configuration to configure the training mode and the superclasses.
Trainer(String outExt, TargetStructure targetStruct, TrainableClassifier[] theClassifiers, Representation theRepresentation, CombinationStrategy combiStrat, TokenizerFactory tFactory, TrainableFilter sentFilter, Set<String> relevantPunct, boolean trainOnlyErrors, boolean testOnly, TiesConfiguration config)
          Creates a new instance.
 



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