Package jsat.clustering.dissimilarity

Interface Summary

Interface	Description
ClusterDissimilarity	This interface provides the basic contract for measuring the dissimilarity between two clusters, and intended for use in Hierarchical Agglomerative Clustering.
UpdatableClusterDissimilarity	This interface extends the contract of a ClusterDissimilarity for more efficient computation.

Class Summary

Class	Description
AbstractClusterDissimilarity	This base class does not currently provide any inheritable functionality, but stores static methods.
AverageLinkDissimilarity	Also known as Group-Average Agglomerative Clustering (GAAC) and UPGMA, this measure computer the dissimilarity by summing the distances between all possible data point pairs in the union of the clusters.
CentroidDissimilarity	Average similarity of all data point pairs between clusters, inter-cluster pairs are ignored.
CompleteLinkDissimilarity	Measures the dissimilarity of two clusters by returning the value of the maximal dissimilarity of any two pairs of data points where one is from each cluster.
DistanceMetricDissimilarity	A base class for Dissimilarity measures that are build ontop the use of some distance metric.
LanceWilliamsDissimilarity	This class provides a base implementation of a Lance Williams (LW) Dissimilarity measure, which is updatable.
MedianDissimilarity	Median link dissimilarity, also called WPGMC.
SingleLinkDissimilarity	Measures the dissimilarity of two clusters by returning the minimum dissimilarity between the two closest data points from the clusters, ie: the minimum distance needed to link the two clusters.
WardsDissimilarity	An implementation of Ward's method for hierarchical clustering.