Package elki.clustering.em.models

Class DiagonalGaussianModelFactory

java.lang.Object

elki.clustering.em.models.DiagonalGaussianModelFactory

All Implemented Interfaces:

EMClusterModelFactory<NumberVector,EMModel>

public class DiagonalGaussianModelFactory
extends java.lang.Object
implements EMClusterModelFactory<NumberVector,EMModel>

Factory for EM with multivariate gaussian models using diagonal matrixes.

These models have individual variances, but no covariance, so this corresponds to the 'VVI' model in Mclust (R).

Since:

0.7.0

Author:

Erich Schubert

Field Summary

Fields

Modifier and Type	Field	Description
protected KMeansInitialization	initializer	Class to choose the initial means

Fields inherited from interface elki.clustering.em.models.EMClusterModelFactory

INIT_ID

Constructor Summary

Constructors

Constructor	Description
DiagonalGaussianModelFactory(KMeansInitialization initializer)	Constructor.

Method Summary

Modifier and Type	Method	Description
java.util.List<DiagonalGaussianModel>	buildInitialModels(Relation<? extends NumberVector> relation, int k)	Build the initial models

Methods inherited from class java.lang.Object

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

Field Detail

initializer

protected KMeansInitialization initializer

Class to choose the initial means

Constructor Detail

DiagonalGaussianModelFactory

public DiagonalGaussianModelFactory(KMeansInitialization initializer)

Constructor.

Parameters:

initializer - Class for choosing the initial seeds.

Method Detail

buildInitialModels

public java.util.List<DiagonalGaussianModel> buildInitialModels(Relation<? extends NumberVector> relation,
                                                                int k)

Description copied from interface: EMClusterModelFactory

Build the initial models

Specified by:

buildInitialModels in interface EMClusterModelFactory<NumberVector,EMModel>

Parameters:

relation - Relation

k - Number of clusters

Returns:

Initial models