smile.stat.distribution

Class ExponentialFamilyMixture

java.lang.Object

smile.stat.distribution.AbstractDistribution

smile.stat.distribution.Mixture

smile.stat.distribution.ExponentialFamilyMixture

All Implemented Interfaces:

Distribution

Direct Known Subclasses:

GaussianMixture

public class ExponentialFamilyMixture
extends Mixture

The finite mixture of distributions from exponential family. The EM algorithm can be used to learn the mixture model from data. EM is particularly useful when the likelihood is an exponential family: the E-step becomes the sum of expectations of sufficient statistics, and the M-step involves maximizing a linear function. In such a case, it is usually possible to derive closed form updates for each step.

Author:

Haifeng Li

Nested Class Summary

Nested classes/interfaces inherited from class smile.stat.distribution.Mixture

Mixture.Component

Constructor Summary

Constructors

Constructor and Description
ExponentialFamilyMixture(List<Mixture.Component> mixture) Constructor.
ExponentialFamilyMixture(List<Mixture.Component> mixture, double[] data) Constructor.

Method Summary

Methods inherited from class smile.stat.distribution.Mixture

bic, cdf, entropy, getComponents, logp, mean, npara, p, quantile, rand, sd, size, toString, var

Methods inherited from class smile.stat.distribution.AbstractDistribution

inverseTransformSampling, likelihood, logLikelihood, quantile, quantile, rejection

Methods inherited from class java.lang.Object

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

Constructor Detail

ExponentialFamilyMixture

public ExponentialFamilyMixture(List<Mixture.Component> mixture)

Constructor.

Parameters:

mixture - a list of exponential family distributions.

ExponentialFamilyMixture

public ExponentialFamilyMixture(List<Mixture.Component> mixture,
                        double[] data)

Constructor. The mixture model will be learned from the given data with the EM algorithm.

Parameters:

mixture - the initial guess of mixture. Components may have different distribution form.

data - the training data.