Examples of cc.mallet.classify.Classifier

cc.mallet.classify.Classifier
Abstract parent of all Classifiers.
All classification techniques in MALLET are implemented as two classes: a trainer and a classifier. The trainer injests the training data and creates a classifier that holds the parameters set during training. The classifier applies those parameters to an Instance to produce a classification of the Instance.
A concrete classifier is required only to be able to classify an instance.
Methods for classifying an InstanceList are here. There are also methods for calculating precison, recall, and f1 from either InstanceLists (which are classified first) or an ArrayList of classifications. Similar functionality is also in {@link cc.mallet.classify.Trial}
A classifier holds a reference to the pipe that was used to create the Instances being classified. Most classifiers use this to make sure the Alphabets of the instances being classified are the same Alphabet objects used during training.
Alphabets are allowed to between training and classification. @see ClassifierTrainer @see Instance @see InstanceList @see Classification @see Trial

        f++;
        InstanceList[] fold = cvIter.nextSplit();


        logger.info("Training token classifier on cv fold " + f + " / " + m_numCV + " (size=" + fold[0].size() + ")...");
        
        Classifier foldClassifier = m_trainer.train(fold[0]);
        Trial t1 = new Trial(foldClassifier, fold[0]);
        Trial t2 = new Trial(foldClassifier, fold[1]);


        logger.info("Within-fold accuracy = " + t1.getAccuracy());
        logger.info("Out-of-fold accuracy = " + t2.getAccuracy());

View Full Code Here

      Object instName = instance.getName();
      
      if (! useOutOfFold || ! m_table.containsKey(instName))
        return m_tokenClassifier.classify(instance);
      
      Classifier classifier = (Classifier) m_table.get(instName);


      return classifier.classify(instance);
    }

View Full Code Here

    Pipe clusterPipe = new OverlappingFeaturePipe();
    System.err.println("Training with " + training);
    InstanceList trainList = new InstanceList(clusterPipe);
    trainList.addThruPipe(new ClusterSampleIterator(training, random, 0.5, 100));
    System.err.println("Created " + trainList.size() + " instances.");
    Classifier me = new MaxEntTrainer().train(trainList);
    ClassifyingNeighborEvaluator eval =
      new ClassifyingNeighborEvaluator(me, "YES");
                                           
    Trial trial = new Trial(me, trainList);
    System.err.println(new ConfusionMatrix(trial));

View Full Code Here

          trainingInstances.add(pipe.pipe(inst));
        }
      }
      logger.info("generated " + trainingInstances.size()
          + " training instances");
      Classifier classifier = new MaxEntTrainer().train(trainingInstances);
      logger.info("InfoGain:\n");
      new InfoGain(trainingInstances).printByRank(System.out);
      logger.info("pairwise training accuracy="
          + new Trial(classifier, trainingInstances).getAccuracy());
      NeighborEvaluator neval = new PairwiseEvaluator(classifier, "YES",

View Full Code Here

TOP

Related Classes of cc.mallet.classify.Classifier

cc.mallet.cluster.examples.FirstOrderClusterExample

cc.mallet.cluster.tui.Clusterings2Clusterer

cc.mallet.pipe.AddClassifierTokenPredictions$TokenClassifiers

cc.mallet.types.Instance

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.