gov.sandia.cognition.statistics.bayesian
Interface DirichletProcessMixtureModel.Updater<ObservationType>

Type Parameters:
ObservationType - Type of observations handled by the mixture model
All Superinterfaces:
Cloneable, CloneableSerializable, Serializable
All Known Implementing Classes:
DirichletProcessMixtureModel.MultivariateMeanCovarianceUpdater, DirichletProcessMixtureModel.MultivariateMeanUpdater
Enclosing class:
DirichletProcessMixtureModel<ObservationType>

public static interface DirichletProcessMixtureModel.Updater<ObservationType>
extends CloneableSerializable

Updater for the DPMM


Method Summary
 ProbabilityFunction<ObservationType> createClusterPosterior(Iterable<? extends ObservationType> values, Random random)
          Updates the cluster from the values assigned to it
 ProbabilityFunction<ObservationType> createPriorPredictive(Iterable<? extends ObservationType> data)
          Creates the prior predictive distribution from the data.
 
Methods inherited from interface gov.sandia.cognition.util.CloneableSerializable
clone
 

Method Detail

createPriorPredictive

ProbabilityFunction<ObservationType> createPriorPredictive(Iterable<? extends ObservationType> data)
Creates the prior predictive distribution from the data.

Parameters:
data - Data from which to create the prior predictive
Returns:
Prior predictive distribution from the data

createClusterPosterior

ProbabilityFunction<ObservationType> createClusterPosterior(Iterable<? extends ObservationType> values,
                                                            Random random)
Updates the cluster from the values assigned to it

Parameters:
values - Values assigned to the cluster
random - Random number generator
Returns:
Updated cluster value