GClasses

GClasses::GBayesianModelAveraging Class Reference

This is an ensemble that uses the bagging approach for training, and Bayesian Model Averaging to combine the models. That is, it trains each model with data drawn randomly with replacement from the original training data. It combines the models with weights proporitional to their likelihood as computed using Bayes' law. More...

#include <GEnsemble.h>

Inheritance diagram for GClasses::GBayesianModelAveraging:
GClasses::GBag GClasses::GEnsemble GClasses::GSupervisedLearner GClasses::GTransducer

List of all members.

Public Member Functions

 GBayesianModelAveraging (GRand &rand)
 General-purpose constructor.
 GBayesianModelAveraging (GDomNode *pNode, GLearnerLoader &ll)
 Deserializing constructor.
virtual ~GBayesianModelAveraging ()
virtual GDomNodeserialize (GDom *pDoc)
 Marshal this object into a DOM, which can then be converted to a variety of serial formats.

Protected Member Functions

virtual bool canImplicitlyHandleContinuousLabels ()
 See the comment for GLearner::canImplicitlyHandleContinuousLabels.
virtual void determineWeights (GMatrix &features, GMatrix &labels)
 Determines the weights in the manner of Bayesian model averaging, with the assumption of uniform priors.

Detailed Description

This is an ensemble that uses the bagging approach for training, and Bayesian Model Averaging to combine the models. That is, it trains each model with data drawn randomly with replacement from the original training data. It combines the models with weights proporitional to their likelihood as computed using Bayes' law.


Constructor & Destructor Documentation

GClasses::GBayesianModelAveraging::GBayesianModelAveraging ( GRand rand) [inline]

General-purpose constructor.

GClasses::GBayesianModelAveraging::GBayesianModelAveraging ( GDomNode pNode,
GLearnerLoader ll 
) [inline]

Deserializing constructor.

virtual GClasses::GBayesianModelAveraging::~GBayesianModelAveraging ( ) [inline, virtual]

Member Function Documentation

virtual bool GClasses::GBayesianModelAveraging::canImplicitlyHandleContinuousLabels ( ) [inline, protected, virtual]

See the comment for GLearner::canImplicitlyHandleContinuousLabels.

Reimplemented from GClasses::GTransducer.

virtual void GClasses::GBayesianModelAveraging::determineWeights ( GMatrix features,
GMatrix labels 
) [protected, virtual]

Determines the weights in the manner of Bayesian model averaging, with the assumption of uniform priors.

Reimplemented from GClasses::GBag.

virtual GDomNode* GClasses::GBayesianModelAveraging::serialize ( GDom pDoc) [virtual]

Marshal this object into a DOM, which can then be converted to a variety of serial formats.

Reimplemented from GClasses::GBag.