marytts.machinelearning

Class GMMTrainer

java.lang.Object

marytts.machinelearning.GMMTrainer

public class GMMTrainer
extends Object

Expectation-Maximization (EM) based GMM training Reference: A. P. Dempster, N. M. Laird, and D. B. Rubin. Maximum likelihood from in- complete data via the em algorithm. Journal of the Royal Statistical Society: Series B, 39(1):1–38, November 1977.

Author:

Oytun Türk

Field Summary

Fields

Modifier and Type	Field and Description
double[]	logLikelihoods

Constructor Summary

Constructors

Constructor and Description
GMMTrainer()

Method Summary

Methods

Modifier and Type	Method and Description
GMM	expectationMaximization(double[][] x, GMM initialGmm, int emMinimumIterations, int emMaximumIterations, boolean isUpdateCovariances, double tinyLogLikelihoodChangePercent, double minimumCovarianceAllowed)
static void	main(String[] args)
static void	testEndianFileIO()
GMM	train(double[][] x, GMMTrainerParams gmmParams)

Methods inherited from class java.lang.Object

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

Field Detail

logLikelihoods

public double[] logLikelihoods

Constructor Detail

GMMTrainer

public GMMTrainer()

Method Detail

train

public GMM train(double[][] x,
        GMMTrainerParams gmmParams)

expectationMaximization

public GMM expectationMaximization(double[][] x,
                          GMM initialGmm,
                          int emMinimumIterations,
                          int emMaximumIterations,
                          boolean isUpdateCovariances,
                          double tinyLogLikelihoodChangePercent,
                          double minimumCovarianceAllowed)

testEndianFileIO

public static void testEndianFileIO()
                             throws IOException

Throws:

IOException

main

public static void main(String[] args)