Package ristretto.jmltools.clustering.evaluation

Class DaviesBouldin

  • java.lang.Object
    • ristretto.jmltools.clustering.evaluation.DaviesBouldin
  • All Implemented Interfaces:
    net.sf.javaml.clustering.evaluation.ClusterEvaluation
    public class DaviesBouldin
extends java.lang.Object
implements net.sf.javaml.clustering.evaluation.ClusterEvaluation
    This class implements the CVI proposed by Davies and Bouldin in [1].

    [1] D. L. Davies and D. W. Bouldin. A clustering separation measure. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1:224-227, 1979. https://doi.org/10.1109/TPAMI.1979.4766909

    • Constructor Summary

      Constructors 
      Constructor Description
      DaviesBouldin()
      Construct a new evaluator that will use the Euclidean distance to measure the errors.
      DaviesBouldin​(net.sf.javaml.distance.DistanceMeasure dm)
      Construct a new evaluator that will use the supplied distance metric to measure the errors

    • Method Summary

      All Methods Instance Methods Concrete Methods 
      Modifier and Type Method Description
      boolean compareScore​(double score1, double score2)
      Compare the two scores according to the criterion in the implementation.
      double score​(net.sf.javaml.core.Dataset[] clusters)
      Return the score the current clusterer obtains on the dataset

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

    • Constructor Detail

      • DaviesBouldin

        public DaviesBouldin()
        Construct a new evaluator that will use the Euclidean distance to measure the errors.

      • DaviesBouldin

        public DaviesBouldin​(net.sf.javaml.distance.DistanceMeasure dm)
        Construct a new evaluator that will use the supplied distance metric to measure the errors
        Parameters:
        dm - Distance measure to be used in the evaluator

    • Method Detail

      • score

        public double score​(net.sf.javaml.core.Dataset[] clusters)
        Return the score the current clusterer obtains on the dataset
        Specified by:
        score in interface net.sf.javaml.clustering.evaluation.ClusterEvaluation
        Parameters:
        clusters - Clusters to be evaluated
        Returns:
        The score the clusterer obtained on this particular dataset

      • compareScore

        public boolean compareScore​(double score1,
                            double score2)
        Compare the two scores according to the criterion in the implementation. Some criterions should be maximized, others should be minimized. This method returns true if the second score is 'better' than the first one
        Specified by:
        compareScore in interface net.sf.javaml.clustering.evaluation.ClusterEvaluation
        Parameters:
        score1 - The first score
        score2 - The second score
        Returns:
        true if the second score is better than the first, false in all other cases