@Title("K-means clustering")

Package elki.clustering.kmeans

K-means clustering and variations.

Interface Summary

Interface	Description
KMeans<V extends NumberVector,M extends Model>	Some constants and options shared among kmeans family algorithms.

Class Summary

Class	Description
AbstractKMeans<V extends NumberVector,M extends Model>	Abstract base class for k-means implementations.
AbstractKMeans.Instance	Inner instance for a run, for better encapsulation, that encapsulates the standard flow of most (but not all) k-means variations.
AbstractKMeans.Par<V extends NumberVector>	Parameterization class.
AnnulusKMeans<V extends NumberVector>	Annulus k-means algorithm.
AnnulusKMeans.Instance	Inner instance, storing state for a single data set.
AnnulusKMeans.Par<V extends NumberVector>	Parameterization class.
BestOfMultipleKMeans<V extends NumberVector,M extends MeanModel>	Run K-Means multiple times, and keep the best run.
BetulaLloydKMeans	BIRCH/BETULA-based clustering algorithm that simply treats the leafs of the CFTree as clusters.
BetulaLloydKMeans.Par	Parameterization class.
BisectingKMeans<V extends NumberVector,M extends MeanModel>	The bisecting k-means algorithm works by starting with an initial partitioning into two clusters, then repeated splitting of the largest cluster to get additional clusters.
CompareMeans<V extends NumberVector>	Compare-Means: Accelerated k-means by exploiting the triangle inequality and pairwise distances of means to prune candidate means.
CompareMeans.Instance	Inner instance, storing state for a single data set.
CompareMeans.Par<V extends NumberVector>	Parameterization class.
ElkanKMeans<V extends NumberVector>	Elkan's fast k-means by exploiting the triangle inequality.
ElkanKMeans.Instance	Inner instance, storing state for a single data set.
ElkanKMeans.Par<V extends NumberVector>	Parameterization class.
ExponionKMeans<V extends NumberVector>	Newlings's Exponion k-means algorithm, exploiting the triangle inequality.
ExponionKMeans.Instance	Inner instance, storing state for a single data set.
ExponionKMeans.Par<V extends NumberVector>	Parameterization class.
FuzzyCMeans<V extends NumberVector>	Fuzzy Clustering developed by Dunn and revisited by Bezdek
FuzzyCMeans.Par	Parameterization class.
GMeans<V extends NumberVector,M extends MeanModel>	G-Means extends K-Means and estimates the number of centers with Anderson Darling Test. Implemented as specialization of XMeans.
GMeans.Par<V extends NumberVector,M extends MeanModel>	Parameterization class.
HamerlyKMeans<V extends NumberVector>	Hamerly's fast k-means by exploiting the triangle inequality.
HamerlyKMeans.Instance	Inner instance, storing state for a single data set.
HamerlyKMeans.Par<V extends NumberVector>	Parameterization class.
HartiganWongKMeans<V extends NumberVector>	Hartigan and Wong k-means clustering.
HartiganWongKMeans.Instance	Instance for a particular data set.
HartiganWongKMeans.Parameterizer<V extends NumberVector>	Parameterization class.
KDTreeFilteringKMeans<V extends NumberVector>	Filtering or "blacklisting" K-means with k-d-tree acceleration.
KDTreeFilteringKMeans.Par<V extends NumberVector>	Parameterization class.
KDTreePruningKMeans<V extends NumberVector>	Pruning K-means with k-d-tree acceleration.
KDTreePruningKMeans.KDNode	Node of the k-d-tree used internally.
KDTreePruningKMeans.Par<V extends NumberVector>	Parameterization class.
KMeansMinusMinus<V extends NumberVector>	k-means--: A Unified Approach to Clustering and Outlier Detection.
KMeansMinusMinus.Par<V extends NumberVector>	Parameterization class.
KMediansLloyd<V extends NumberVector>	k-medians clustering algorithm, but using Lloyd-style bulk iterations instead of the more complicated approach suggested by Kaufman and Rousseeuw (see PAM instead).
KMediansLloyd.Instance	Inner instance, storing state for a single data set.
KMediansLloyd.Par<V extends NumberVector>	Parameterization class.
LloydKMeans<V extends NumberVector>	The standard k-means algorithm, using bulk iterations and commonly attributed to Lloyd and Forgy (independently).
LloydKMeans.Instance	Inner instance, storing state for a single data set.
LloydKMeans.Par<V extends NumberVector>	Parameterization class.
MacQueenKMeans<V extends NumberVector>	The original k-means algorithm, using MacQueen style incremental updates; making this effectively an "online" (streaming) algorithm.
MacQueenKMeans.Instance	Inner instance, storing state for a single data set.
MacQueenKMeans.Par<V extends NumberVector>	Parameterization class.
ShallotKMeans<V extends NumberVector>	Borgelt's Shallot k-means algorithm, exploiting the triangle inequality.
ShallotKMeans.Instance	Inner instance, storing state for a single data set.
ShallotKMeans.Par<V extends NumberVector>	Parameterization class.
SimplifiedElkanKMeans<V extends NumberVector>	Simplified version of Elkan's k-means by exploiting the triangle inequality.
SimplifiedElkanKMeans.Instance	Inner instance, storing state for a single data set.
SimplifiedElkanKMeans.Par<V extends NumberVector>	Parameterization class.
SingleAssignmentKMeans<V extends NumberVector>	Pseudo-k-means variations, that assigns each object to the nearest center.
SingleAssignmentKMeans.Instance	Inner instance, storing state for a single data set.
SingleAssignmentKMeans.Par<V extends NumberVector>	Parameterization class.
SortMeans<V extends NumberVector>	Sort-Means: Accelerated k-means by exploiting the triangle inequality and pairwise distances of means to prune candidate means (with sorting).
SortMeans.Instance	Inner instance, storing state for a single data set.
SortMeans.Par<V extends NumberVector>	Parameterization class.
XMeans<V extends NumberVector,M extends MeanModel>	X-means: Extending K-means with Efficient Estimation on the Number of Clusters.
YinYangKMeans<V extends NumberVector>	Yin-Yang k-Means Clustering.
YinYangKMeans.Instance	Instance for a particular data set.
YinYangKMeans.Par<V extends NumberVector>	Parameterization class.

Enum Summary

Enum	Description
KDTreePruningKMeans.Split	Splitting strategies for constructing the k-d-tree.