Package | Description |
---|---|
messif.buckets |
Storage classes capable of holding
objects . |
messif.pivotselection |
Support for selection of representative objects (pivots).
|
messif.utility |
Various utilities that does not fit anywhere else including
a main class for executing batch files.
|
Class and Description |
---|
AbstractPivotChooser
Abstract class for pivot selection algorithms hierarchy
This class provides basic methods for selecting and accessing pivots and automatically registers
statistic DistanceComputations.PivotChooser, i.e. number of distance computations
spent in choosing pivots.
|
Class and Description |
---|
AbstractPivotChooser
Abstract class for pivot selection algorithms hierarchy
This class provides basic methods for selecting and accessing pivots and automatically registers
statistic DistanceComputations.PivotChooser, i.e. number of distance computations
spent in choosing pivots.
|
ClusterPivotChooser.Cluster
Class encapsulating objects of one cluster and storing the cluster's radius.
|
ClusterPivotChooser.Pair
Class encapsulating two clusters and the diameters of a cluster that would be produced be merging these clusters.
|
ClusterPivotChooser.PrecomputedDistances
A cache for distances between a pair of objects
|
CoveragePivotChooser.Ball
Class encapsulating info about each cluster (ball region).
|
CoveragePivotChooser.PrecomputedDistances
A cache for distances between a pair of objects
|
KMeansPivotChooser
This class uses the k-means algorithm adapted for metric spaces to cluster the objects,
so it is k-medoids algorithm in fact.
|
KMeansPivotChooser.CenterThread
Internal abract thread for selecting new "center" of a cluster.
|
RandomPivotChooser
RandomPivotChooser provides the capability of selecting a random object from the whole bucket.
|
Class and Description |
---|
AbstractPivotChooser
Abstract class for pivot selection algorithms hierarchy
This class provides basic methods for selecting and accessing pivots and automatically registers
statistic DistanceComputations.PivotChooser, i.e. number of distance computations
spent in choosing pivots.
|