Classes for training and classifying instances. 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.
Interface Summary Interface Description BoostableThis interface is a tag indicating that the classifier attends to the InstanceList.getInstanceWeight() weights when training. ClassifierTrainer.ByActiveLearning<C extends Classifier>For active learning, in which this trainer will select certain instances and request that the Labeler instance label them. ClassifierTrainer.ByIncrements<C extends Classifier>For various kinds of online learning by batches, where training instances are presented, consumed for learning immediately. ClassifierTrainer.ByInstanceIncrements<C extends Classifier>For online learning that can operate on one instance at a time. ClassifierTrainer.ByOptimization<C extends Classifier>
Class Summary Class Description AdaBoostAdaBoost Robert E. AdaBoostM2AdaBoostM2 AdaBoostM2TrainerThis version of AdaBoost can handle multi-class problems. AdaBoostTrainerThis version of AdaBoost should be used only for binary classification. BaggingClassifier BaggingTrainerBagging Trainer. BalancedWinnowClassification methods of BalancedWinnow algorithm. BalancedWinnowTrainerAn implementation of the training methods of a BalancedWinnow on-line classifier. C45A C4.5 Decision Tree classifier. C45.Node C45TrainerA C4.5 decision tree learner, approximtely. ClassificationThe result of classifying a single instance. ClassifierAbstract parent of all Classifiers. ClassifierAccuracyEvaluator ClassifierEnsembleClassifer for an ensemble of classifers, combined with learned weights. ClassifierEnsembleTrainer ClassifierEvaluator ClassifierTrainer<C extends Classifier>Each ClassifierTrainer trains one Classifier based on various interfaces for consuming training data. ClassifierTrainer.Factory<CT extends ClassifierTrainer<? extends Classifier>>Instances of a Factory know how to create new ClassifierTrainers to apply to new Classifiers. ConfidencePredictingClassifier ConfidencePredictingClassifierTrainer DecisionTreeDecision Tree classifier. DecisionTree.Node DecisionTreeTrainerA decision tree learner, roughly ID3, but only to a fixed given depth in all branches. DecisionTreeTrainer.Factory FeatureConstraintUtilUtility functions for creating feature constraints that can be used with GE training. FeatureSelectingClassifierTrainerAdaptor for adding feature selection to a classifier trainer. MaxEntMaximum Entropy (AKA Multivariate Logistic Regression) classifier. MaxEntGERangeTrainerTraining of MaxEnt models with labeled features using Generalized Expectation Criteria. MaxEntGETrainerTraining of MaxEnt models with labeled features using Generalized Expectation Criteria. MaxEntL1Trainer MaxEntOptimizableByGE MaxEntOptimizableByLabelDistribution MaxEntOptimizableByLabelLikelihood MaxEntPRTrainerPenalty (soft) version of Posterior Regularization (PR) for training MaxEnt. MaxEntTrainerThe trainer for a Maximum Entropy classifier. MCMaxEntMaximum Entropy classifier. MCMaxEntTrainerThe trainer for a Maximum Entropy classifier. MostFrequentClassAssignmentTrainerA Classifier Trainer to be used with MostFrequentClassifier. MostFrequentClassifierA Classifier that will return the most frequent class label based on a training set. NaiveBayesA classifier that classifies instances according to the NaiveBayes method. NaiveBayesEMTrainer NaiveBayesTrainerClass used to generate a NaiveBayes classifier from a set of training data. NaiveBayesTrainer.Factory PRAuxClassifierAuxiliary model (q) for E-step/I-projection in PR training. PRAuxClassifierOptimizableOptimizable for training auxiliary model (q) for E-step/I-projection in PR training. RandomAssignmentTrainerA Classifier Trainer to be used with RandomClassifier. RandomClassifierA Classifier that will return a randomly selected class label. RankMaxEntRank Maximum Entropy classifier. RankMaxEntTrainerThe trainer for a
TrialStores the results of classifying a collection of Instances, and provides many methods for evaluating the results. WinnowClassification methods of Winnow2 algorithm. WinnowTrainerAn implementation of the training methods of a Winnow2 on-line classifier.