36#ifndef STK_IMIXTURECOMPOSER_H
37#define STK_IMIXTURECOMPOSER_H
142 virtual void pStep();
In this file we define the abstract base class for mixture statistical models.
Base class for Mixture (composed) model.
virtual ~IMixtureComposer()
destructor
virtual int cStep()
Replace tik by zik.
virtual void initializeStep()
Initialize the model before its first use.
void setState(Clust::modelState state)
set the state of the model : should be used by any strategy
void randomFuzzyInit()
Initialize randomly the posterior probabilities tik of the model, then compute the zi values with map...
virtual void pStep()
Compute proportions using the ML estimates, default implementation.
virtual IMixtureComposer * clone() const =0
clone pattern
virtual IMixtureComposer * create() const =0
create pattern
virtual Real eStep()
compute the zi, the lnLikelihood of the current estimates and the next value of the tik.
CPointX lnComp_
Auxiliary array used in the eStep.
virtual int randomZi()
generate random zi_
Clust::modelState state_
state of the model
virtual void initializeMixtureParameters()
Create the mixture model parameters pk_ and tik_.
virtual void mapStep()
Compute zi using the Map estimate.
Clust::modelState state() const
void randomClassInit()
Initialize randomly the labels zi of the model.
virtual int sStep()
Simulate zi accordingly to tik and replace tik by zik by calling cStep().
virtual int randomTik()
generate random tik_
virtual void randomInit()=0
initialize randomly the parameters of the components of the model
virtual void paramUpdateStep()=0
Compute the proportions and the model parameters given the current tik mixture parameters.
virtual void finalizeStep()
Finalize the estimation of the model.
Interface base class for Mixture (composed) model.
The MultidimRegression class allows to regress a multidimensional output variable among a multivariat...
modelState
Give the state of the model.
double Real
STK fundamental type of Real values.
The namespace STK is the main domain space of the Statistical ToolKit project.