org.ddogleg.clustering

Class FactoryClustering

java.lang.Object

org.ddogleg.clustering.FactoryClustering

public class FactoryClustering
extends java.lang.Object

Factory for creating clustering algorithms.

Constructor Summary

Constructors

Constructor and Description
FactoryClustering()

Method Summary

Modifier and Type	Method and Description
static ExpectationMaximizationGmm_F64	gaussianMixtureModelEM_F64(int maxIterations, int maxConverge, double convergeTol) High level interface for creating GMM cluster.
static StandardKMeans_F64	kMeans_F64(KMeansInitializers initializer, int maxIterations, int maxConverge, double convergeTol) High level interface for creating k-means cluster.

Methods inherited from class java.lang.Object

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

Constructor Detail

FactoryClustering

public FactoryClustering()

Method Detail

gaussianMixtureModelEM_F64

public static ExpectationMaximizationGmm_F64 gaussianMixtureModelEM_F64(int maxIterations,
                                                                        int maxConverge,
                                                                        double convergeTol)

High level interface for creating GMM cluster. If more flexibility is needed (e.g. custom seeds) then create and instance of ExpectationMaximizationGmm_F64 directly

WARNING: DEVELOPMENTAL AND IS LIKELY TO FAIL HORRIBLY

Parameters:

maxIterations - Maximum number of iterations it will perform.

maxConverge - Maximum iterations allowed before convergence. Re-seeded if it doesn't converge.

convergeTol - Distance based convergence tolerance. Try 1e-8

Returns:

ExpectationMaximizationGmm_F64

kMeans_F64

public static StandardKMeans_F64 kMeans_F64(KMeansInitializers initializer,
                                            int maxIterations,
                                            int maxConverge,
                                            double convergeTol)

High level interface for creating k-means cluster. If more flexibility is needed (e.g. custom seeds) then create and instance of StandardKMeans_F64 directly

Parameters:

initializer - Specify which method should be used to select the initial seeds for the clusters. null means default.

maxIterations - Maximum number of iterations it will perform.

maxConverge - Maximum iterations allowed before convergence. Re-seeded if it doesn't converge.

convergeTol - Distance based convergence tolerance. Try 1e-8

Returns:

StandardKMeans_F64