Package de.lmu.ifi.dbs.elki.algorithm.clustering

Clustering algorithms.

See: Description

Interface Summary

Interface	Description
ClusteringAlgorithm<C extends Clustering<? extends Model>>	Interface for Algorithms that are capable to provide a Clustering as Result. in general, clustering algorithms are supposed to implement the Algorithm-Interface.
GriDBSCAN.Assignment	Point assignment.

Class Summary

Class	Description
AbstractProjectedClustering<R extends Clustering<?>,V extends NumberVector>	Abstract superclass for projected clustering algorithms, like PROCLUS and ORCLUS.
AbstractProjectedClustering.Parameterizer	Parameterization class.
CanopyPreClustering<O>	Canopy pre-clustering is a simple preprocessing step for clustering.
CanopyPreClustering.Parameterizer<O>	Parameterization class
ClusteringAlgorithmUtil	Utility functionality for writing clustering algorithms.
DBSCAN<O>	Density-Based Clustering of Applications with Noise (DBSCAN), an algorithm to find density-connected sets in a database.
DBSCAN.Parameterizer<O>	Parameterization class.
GriDBSCAN<V extends NumberVector>	Using Grid for Accelerating Density-Based Clustering.
GriDBSCAN.Border	Border point assignment.
GriDBSCAN.Core	Core point assignment.
GriDBSCAN.Instance<V extends NumberVector>	Instance, for a single run.
GriDBSCAN.MultiBorder	Multiple border point assignment.
GriDBSCAN.Parameterizer<O extends NumberVector>	Parameterization class.
NaiveMeanShiftClustering<V extends NumberVector>	Mean-shift based clustering algorithm.
NaiveMeanShiftClustering.Parameterizer<V extends NumberVector>	Parameterizer.
SNNClustering<O>	Shared nearest neighbor clustering.
SNNClustering.Parameterizer<O>	Parameterization class.

Package de.lmu.ifi.dbs.elki.algorithm.clustering Description

Clustering algorithms.

Clustering algorithms are supposed to implement the Algorithm-Interface. The more specialized interface ClusteringAlgorithm requires an implementing algorithm to provide a special result class suitable as a partitioning of the database. More relaxed clustering algorithms are allowed to provide a result that is a fuzzy clustering, does not partition the database complete or is in any other sense a relaxed clustering result.

See Also:

de.lmu.ifi.dbs.elki.algorithm