edu.umd.cloud9.example.clustering
Class ExpectationMaximization

java.lang.Object
  edu.umd.cloud9.example.clustering.ExpectationMaximization

public class ExpectationMaximization
extends Object

Constructor Summary

ExpectationMaximization()

Method Summary

static UnivariateGaussianMixtureModel	initialize(Point[] points, double[] means) Initializes the mixture model with points that are closet the given means.
static UnivariateGaussianMixtureModel	initialize(Point[] points, int n) Initializes the mixture model with random points.
static UnivariateGaussianMixtureModel	run(Point[] points, UnivariateGaussianMixtureModel m) Performs the Expectation-Maximization algorithm.
static Integer[]	sampleNUniquePoints(int n, int length)

Methods inherited from class java.lang.Object

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

Constructor Detail

ExpectationMaximization

public ExpectationMaximization()

Method Detail

initialize

public static UnivariateGaussianMixtureModel initialize(Point[] points,
                                                        double[] means)

Initializes the mixture model with points that are closet the given means.

initialize

public static UnivariateGaussianMixtureModel initialize(Point[] points,
                                                        int n)

Initializes the mixture model with random points.

run

public static UnivariateGaussianMixtureModel run(Point[] points,
                                                 UnivariateGaussianMixtureModel m)

Performs the Expectation-Maximization algorithm. The parameters estimated corresponds to univariate Gaussian distributions.

Parameters:

points - point set

m - initial mixture model

Returns:

learned mixture model

sampleNUniquePoints

public static final Integer[] sampleNUniquePoints(int n,
                                                  int length)