Class HeterogeneousEnsembleBlast

  • All Implemented Interfaces:
    Configurable, Serializable, CapabilitiesHandler, Classifier, MultiClassClassifier, AWTRenderable, Learner<Example<Instance>>, MOAObject, OptionHandler

    public class HeterogeneousEnsembleBlast
    extends HeterogeneousEnsembleAbstract
    implements MultiClassClassifier
    BLAST (Best Last) for Heterogeneous Ensembles implemented with Fading Factors

    Given a set of (heterogeneous) classifiers, BLAST builds an ensemble, and determines the weights of all ensemble members based on their performance on recent observed instances. This implementation uses a window of recent instances, the size is determined by the 'w' parameter.

    J. N. van Rijn, G. Holmes, B. Pfahringer, J. Vanschoren. Having a Blast: Meta-Learning and Heterogeneous Ensembles for Data Streams. In 2015 IEEE International Conference on Data Mining, pages 1003-1008. IEEE, 2015.

    Parameters:

    • -w : Window size
    • -b : Comma-separated string of classifiers
    • -g : Grace period (1 = optimal)
    • -k : Number of active classifiers
    Version:
    $Revision: 1 $
    Author:
    Jan N. van Rijn (j.n.van.rijn@liacs.leidenuniv.nl)
    See Also:
    Serialized Form
    • Field Detail

      • onlineHistory

        protected boolean[][] onlineHistory
      • windowSizeOption

        public IntOption windowSizeOption
    • Constructor Detail

      • HeterogeneousEnsembleBlast

        public HeterogeneousEnsembleBlast()
    • Method Detail

      • resetLearningImpl

        public void resetLearningImpl()
        Description copied from class: AbstractClassifier
        Resets this classifier. It must be similar to starting a new classifier from scratch.

        The reason for ...Impl methods: ease programmer burden by not requiring them to remember calls to super in overridden methods. Note that this will produce compiler errors if not overridden.
        Specified by:
        resetLearningImpl in class AbstractClassifier
      • trainOnInstanceImpl

        public void trainOnInstanceImpl​(Instance inst)
        Description copied from class: AbstractClassifier
        Trains this classifier incrementally using the given instance.

        The reason for ...Impl methods: ease programmer burden by not requiring them to remember calls to super in overridden methods. Note that this will produce compiler errors if not overridden.
        Specified by:
        trainOnInstanceImpl in class AbstractClassifier
        Parameters:
        inst - the instance to be used for training