GestureRecognitionToolkit  Version: 0.2.5
The Gesture Recognition Toolkit (GRT) is a cross-platform, open-source, c++ machine learning library for real-time gesture recognition.
Public Member Functions | Static Public Member Functions | Protected Member Functions | Protected Attributes | List of all members
GaussianMixtureModels Class Reference

#include <GaussianMixtureModels.h>

Inheritance diagram for GaussianMixtureModels:
Clusterer MLBase GRTBase Observer< TrainingResult > Observer< TestInstanceResult >

Public Member Functions

 GaussianMixtureModels (const UINT numClusters=10, const UINT minNumEpochs=5, const UINT maxNumEpochs=1000, const Float minChange=1.0e-5, const UINT numRestarts=5)
 
 GaussianMixtureModels (const GaussianMixtureModels &rhs)
 
virtual ~GaussianMixtureModels ()
 
GaussianMixtureModels & operator= (const GaussianMixtureModels &rhs)
 
virtual bool deepCopyFrom (const Clusterer *clusterer)
 
virtual bool reset ()
 
virtual bool clear ()
 
virtual bool train_ (MatrixFloat &trainingData)
 
virtual bool train_ (ClassificationData &trainingData)
 
virtual bool train_ (UnlabelledData &trainingData)
 
virtual bool predict_ (VectorDouble &inputVector)
 
virtual bool saveModelToFile (std::fstream &file) const
 
virtual bool loadModelFromFile (std::fstream &file)
 
MatrixFloat getMu () const
 
Vector< MatrixFloat > getSigma () const
 
MatrixFloat getSigma (const UINT k) const
 
bool setNumRestarts (const UINT numRestarts)
 
- Public Member Functions inherited from Clusterer
 Clusterer (const std::string &id="")
 
virtual ~Clusterer (void)
 
bool copyBaseVariables (const Clusterer *clusterer)
 
UINT getNumClusters () const
 
UINT getPredictedClusterLabel () const
 
Float getMaximumLikelihood () const
 
Float getBestDistance () const
 
VectorFloat getClusterLikelihoods () const
 
VectorFloat getClusterDistances () const
 
Vector< UINT > getClusterLabels () const
 
 GRT_DEPRECATED_MSG ("getClustererType() is deprecated, use getId() or getBaseId() instead", std::string getClustererType() const )
 
bool setNumClusters (const UINT numClusters)
 
Clusterer * create () const
 
 GRT_DEPRECATED_MSG ("createNewInstance is deprecated, use create instead.", Clusterer *createNewInstance() const )
 
 GRT_DEPRECATED_MSG ("createInstanceFromString is deprecated, use create instead.", static Clusterer *createInstanceFromString(const std::string &id))
 
Clusterer * deepCopy () const
 
const Clusterer & getBaseClusterer () const
 
- Public Member Functions inherited from MLBase
 MLBase (const std::string &id="", const BaseType type=BASE_TYPE_NOT_SET)
 
virtual ~MLBase (void)
 
bool copyMLBaseVariables (const MLBase *mlBase)
 
virtual bool train (ClassificationData trainingData)
 
virtual bool train (RegressionData trainingData)
 
virtual bool train_ (RegressionData &trainingData)
 
virtual bool train (RegressionData trainingData, RegressionData validationData)
 
virtual bool train_ (RegressionData &trainingData, RegressionData &validationData)
 
virtual bool train (TimeSeriesClassificationData trainingData)
 
virtual bool train_ (TimeSeriesClassificationData &trainingData)
 
virtual bool train (ClassificationDataStream trainingData)
 
virtual bool train_ (ClassificationDataStream &trainingData)
 
virtual bool train (UnlabelledData trainingData)
 
virtual bool train (MatrixFloat data)
 
virtual bool predict (VectorFloat inputVector)
 
virtual bool predict (MatrixFloat inputMatrix)
 
virtual bool predict_ (MatrixFloat &inputMatrix)
 
virtual bool map (VectorFloat inputVector)
 
virtual bool map_ (VectorFloat &inputVector)
 
virtual bool print () const
 
virtual bool save (const std::string &filename) const
 
virtual bool load (const std::string &filename)
 
virtual bool save (std::fstream &file) const
 
virtual bool load (std::fstream &file)
 
 GRT_DEPRECATED_MSG ("saveModelToFile(std::string filename) is deprecated, use save(const std::string &filename) instead", virtual bool saveModelToFile(const std::string &filename) const )
 
 GRT_DEPRECATED_MSG ("saveModelToFile(std::fstream &file) is deprecated, use save(std::fstream &file) instead", virtual bool saveModelToFile(std::fstream &file) const )
 
 GRT_DEPRECATED_MSG ("loadModelFromFile(std::string filename) is deprecated, use load(const std::string &filename) instead", virtual bool loadModelFromFile(const std::string &filename))
 
 GRT_DEPRECATED_MSG ("loadModelFromFile(std::fstream &file) is deprecated, use load(std::fstream &file) instead", virtual bool loadModelFromFile(std::fstream &file))
 
virtual bool getModel (std::ostream &stream) const
 
virtual std::string getModelAsString () const
 
DataType getInputType () const
 
DataType getOutputType () const
 
BaseType getType () const
 
UINT getNumInputFeatures () const
 
UINT getNumInputDimensions () const
 
UINT getNumOutputDimensions () const
 
UINT getMinNumEpochs () const
 
UINT getMaxNumEpochs () const
 
UINT getBatchSize () const
 
UINT getNumRestarts () const
 
UINT getValidationSetSize () const
 
UINT getNumTrainingIterationsToConverge () const
 
Float getMinChange () const
 
Float getLearningRate () const
 
Float getRMSTrainingError () const
 
 GRT_DEPRECATED_MSG ("getRootMeanSquaredTrainingError() is deprecated, use getRMSTrainingError() instead", Float getRootMeanSquaredTrainingError() const )
 
Float getTotalSquaredTrainingError () const
 
Float getRMSValidationError () const
 
Float getValidationSetAccuracy () const
 
VectorFloat getValidationSetPrecision () const
 
VectorFloat getValidationSetRecall () const
 
bool getUseValidationSet () const
 
bool getRandomiseTrainingOrder () const
 
bool getTrained () const
 
 GRT_DEPRECATED_MSG ("getModelTrained() is deprecated, use getTrained() instead", bool getModelTrained() const )
 
bool getConverged () const
 
bool getScalingEnabled () const
 
bool getIsBaseTypeClassifier () const
 
bool getIsBaseTypeRegressifier () const
 
bool getIsBaseTypeClusterer () const
 
bool getTrainingLoggingEnabled () const
 
bool getTestingLoggingEnabled () const
 
bool enableScaling (const bool useScaling)
 
bool setMaxNumEpochs (const UINT maxNumEpochs)
 
bool setBatchSize (const UINT batchSize)
 
bool setMinNumEpochs (const UINT minNumEpochs)
 
bool setNumRestarts (const UINT numRestarts)
 
bool setMinChange (const Float minChange)
 
bool setLearningRate (const Float learningRate)
 
bool setUseValidationSet (const bool useValidationSet)
 
bool setValidationSetSize (const UINT validationSetSize)
 
bool setRandomiseTrainingOrder (const bool randomiseTrainingOrder)
 
bool setTrainingLoggingEnabled (const bool loggingEnabled)
 
bool setTestingLoggingEnabled (const bool loggingEnabled)
 
bool registerTrainingResultsObserver (Observer< TrainingResult > &observer)
 
bool registerTestResultsObserver (Observer< TestInstanceResult > &observer)
 
bool removeTrainingResultsObserver (const Observer< TrainingResult > &observer)
 
bool removeTestResultsObserver (const Observer< TestInstanceResult > &observer)
 
bool removeAllTrainingObservers ()
 
bool removeAllTestObservers ()
 
bool notifyTrainingResultsObservers (const TrainingResult &data)
 
bool notifyTestResultsObservers (const TestInstanceResult &data)
 
MLBase * getMLBasePointer ()
 
const MLBase * getMLBasePointer () const
 
Vector< TrainingResult > getTrainingResults () const
 
- Public Member Functions inherited from GRTBase
 GRTBase (const std::string &id="")
 
virtual ~GRTBase (void)
 
bool copyGRTBaseVariables (const GRTBase *GRTBase)
 
 GRT_DEPRECATED_MSG ("getClassType is deprecated, use getId() instead!", std::string getClassType() const )
 
std::string getId () const
 
std::string getLastWarningMessage () const
 
std::string getLastErrorMessage () const
 
std::string getLastInfoMessage () const
 
bool setInfoLoggingEnabled (const bool loggingEnabled)
 
bool setWarningLoggingEnabled (const bool loggingEnabled)
 
bool setErrorLoggingEnabled (const bool loggingEnabled)
 
bool setDebugLoggingEnabled (const bool loggingEnabled)
 
GRTBase * getGRTBasePointer ()
 
const GRTBase * getGRTBasePointer () const
 
Float scale (const Float &x, const Float &minSource, const Float &maxSource, const Float &minTarget, const Float &maxTarget, const bool constrain=false)
 
Float SQR (const Float &x) const
 
- Public Member Functions inherited from Observer< TrainingResult >
virtual void notify (const TrainingResult &data)
 
- Public Member Functions inherited from Observer< TestInstanceResult >
virtual void notify (const TestInstanceResult &data)
 

Static Public Member Functions

static std::string getId ()
 
- Static Public Member Functions inherited from Clusterer
static Clusterer * create (std::string const &id)
 
static Vector< std::string > getRegisteredClusterers ()
 
- Static Public Member Functions inherited from GRTBase
static std::string getGRTVersion (bool returnRevision=true)
 
static std::string getGRTRevison ()
 

Protected Member Functions

bool train_ (const UINT numTrainingSamples, const MatrixFloat &data)
 
bool estep (const MatrixFloat &data, VectorDouble &u, VectorDouble &v, Float &change)
 
bool mstep (const MatrixFloat &data)
 
bool computeInvAndDet ()
 
void SWAP (UINT &a, UINT &b)
 
Float SQR (const Float v)
 
Float gauss (const VectorDouble &x, const UINT clusterIndex, const VectorDouble &det, const MatrixFloat &mu, const Vector< MatrixFloat > &invSigma)
 
- Protected Member Functions inherited from Clusterer
bool saveClustererSettingsToFile (std::fstream &file) const
 
bool loadClustererSettingsFromFile (std::fstream &file)
 
- Protected Member Functions inherited from MLBase
bool saveBaseSettingsToFile (std::fstream &file) const
 
bool loadBaseSettingsFromFile (std::fstream &file)
 

Protected Attributes

UINT numTrainingSamples
 The number of samples in the training data.
 
UINT numRestarts
 The number of times the learning algorithm can reattempt to train a model.
 
Float loglike
 The current loglikelihood value of the models given the data.
 
MatrixFloat mu
 A matrix holding the estimated mean values of each Gaussian.
 
MatrixFloat resp
 The responsibility matrix.
 
VectorDouble frac
 A vector holding the P(k)'s.
 
VectorDouble lndets
 A vector holding the log detminants of SIGMA'k.
 
VectorDouble det
 
Vector< MatrixFloat > sigma
 
Vector< MatrixFloat > invSigma
 
- Protected Attributes inherited from Clusterer
UINT numClusters
 Number of clusters in the model.
 
UINT predictedClusterLabel
 Stores the predicted cluster label from the most recent predict( )
 
Float maxLikelihood
 
Float bestDistance
 
VectorFloat clusterLikelihoods
 
VectorFloat clusterDistances
 
Vector< UINT > clusterLabels
 
bool converged
 
Vector< MinMax > ranges
 
- Protected Attributes inherited from MLBase
bool trained
 
bool useScaling
 
bool converged
 
DataType inputType
 
DataType outputType
 
BaseType baseType
 
UINT numInputDimensions
 
UINT numOutputDimensions
 
UINT numTrainingIterationsToConverge
 
UINT minNumEpochs
 
UINT maxNumEpochs
 
UINT batchSize
 
UINT validationSetSize
 
UINT numRestarts
 
Float learningRate
 
Float minChange
 
Float rmsTrainingError
 
Float rmsValidationError
 
Float totalSquaredTrainingError
 
Float validationSetAccuracy
 
bool useValidationSet
 
bool randomiseTrainingOrder
 
VectorFloat validationSetPrecision
 
VectorFloat validationSetRecall
 
Random random
 
Vector< TrainingResult > trainingResults
 
TrainingResultsObserverManager trainingResultsObserverManager
 
TestResultsObserverManager testResultsObserverManager
 
TrainingLog trainingLog
 
TestingLog testingLog
 
- Protected Attributes inherited from GRTBase
std::string classId
 Stores the name of the class (e.g., MinDist)
 
DebugLog debugLog
 
ErrorLog errorLog
 
InfoLog infoLog
 
WarningLog warningLog
 

Additional Inherited Members

- Public Types inherited from Clusterer
typedef std::map< std::string, Clusterer *(*)() > StringClustererMap
 
- Public Types inherited from MLBase
enum  BaseType {
  BASE_TYPE_NOT_SET =0, CLASSIFIER, REGRESSIFIER, CLUSTERER,
  PRE_PROCSSING, POST_PROCESSING, FEATURE_EXTRACTION, CONTEXT
}
 
- Static Protected Member Functions inherited from Clusterer
static StringClustererMap * getMap ()
 

Detailed Description

GRT MIT License Copyright (c) <2012> <Nicholas Gillian, Media Lab, MIT>

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Examples:
ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 37 of file GaussianMixtureModels.h.

Constructor & Destructor Documentation

GaussianMixtureModels::GaussianMixtureModels ( const UINT  numClusters = 10,
const UINT  minNumEpochs = 5,
const UINT  maxNumEpochs = 1000,
const Float  minChange = 1.0e-5,
const UINT  numRestarts = 5 
)

Default Constructor.

Definition at line 15 of file GaussianMixtureModels.cpp.

GaussianMixtureModels::GaussianMixtureModels ( const GaussianMixtureModels &  rhs)

Defines how the data from the rhs instance should be copied to this instance

Parameters
rhsanother instance of a GaussianMixtureModels

Definition at line 26 of file GaussianMixtureModels.cpp.

GaussianMixtureModels::~GaussianMixtureModels ( )
virtual

Default Destructor

Definition at line 46 of file GaussianMixtureModels.cpp.

Member Function Documentation

bool GaussianMixtureModels::clear ( )
virtual

This function clears the Clusterer module, removing any trained model and setting all the base variables to their default values.

Returns
returns true if the derived class was cleared succesfully, false otherwise

Reimplemented from Clusterer.

Definition at line 106 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::deepCopyFrom ( const Clusterer *  clusterer)
virtual

This deep copies the variables and models from the Clusterer pointer to this GaussianMixtureModels instance. This overrides the base deep copy function for the Clusterer modules.

Parameters
clusterera pointer to the Clusterer base class, this should be pointing to another GaussianMixtureModels instance
Returns
returns true if the clone was successfull, false otherwise

Reimplemented from Clusterer.

Definition at line 71 of file GaussianMixtureModels.cpp.

std::string GaussianMixtureModels::getId ( )
static

Gets a string that represents the GaussianMixtureModels class.

Returns
returns a string containing the ID of this class

Definition at line 9 of file GaussianMixtureModels.cpp.

MatrixFloat GaussianMixtureModels::getMu ( ) const
inline

This function returns the mu matrix which is built during the training phase. If the GMM model has not been trained, then this function will return an empty MatrixFloat. If the GMM model has been trained, then each row in the mu matrix represents a cluster and each column represents an input dimension.

Returns
returns the mu matrix if the model has been trained, otherwise an empty MatrixFloat will be returned
Examples:
ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 141 of file GaussianMixtureModels.h.

Vector< MatrixFloat > GaussianMixtureModels::getSigma ( ) const
inline

This function returns the sigma matrix which is built during the training phase. If the GMM model has not been trained, then this function will return an empty vector< MatrixFloat >. If the GMM model has been trained, then each element in the returned vector represents a cluster. Each element is a MatrixFloat, which will have N rows and N columns, where N is the number of input dimensions to the model.

Returns
returns the sigma matrix if the model has been trained, otherwise an empty vector< MatrixFloat > will be returned
Examples:
ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 151 of file GaussianMixtureModels.h.

MatrixFloat GaussianMixtureModels::getSigma ( const UINT  k) const

This function returns the sigma matrix for a specific cluster. If the GMM model has not been trained, then this function will return an empty MatrixFloat. If the GMM model has been trained, then the returned MatrixFloat will have N rows and N columns, where N is the number of input dimensions to the model.

Returns
returns the sigma matrix for a specific cluster if the model has been trained, otherwise an empty MatrixFloat will be returned

Definition at line 551 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::loadModelFromFile ( std::fstream &  file)
virtual

This loads a trained GaussianMixtureModels model from a file. This overrides the loadModelFromFile function in the base class.

Parameters
fstream&file: a reference to the file the GaussianMixtureModels model will be loaded from
Returns
returns true if the model was loaded successfully, false otherwise

Definition at line 366 of file GaussianMixtureModels.cpp.

GaussianMixtureModels & GaussianMixtureModels::operator= ( const GaussianMixtureModels &  rhs)

Defines how the data from the rhs instance should be copied to this instance

Parameters
rhsanother instance of a GaussianMixtureModels
Returns
returns a reference to this instance of the GaussianMixtureModels

Definition at line 49 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::predict_ ( VectorDouble &  inputVector)
virtual

This is the main prediction interface for all the GRT machine learning algorithms. This should be overwritten by the derived class.

Parameters
inputVectora reference to the input vector for prediction
Returns
returns true if the prediction was completed succesfully, false otherwise (the base class always returns false)

Reimplemented from MLBase.

Definition at line 260 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::reset ( )
virtual

This resets the Clusterer. This overrides the reset function in the MLBase base class.

Returns
returns true if the Clusterer was reset, false otherwise

Reimplemented from Clusterer.

Definition at line 96 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::saveModelToFile ( std::fstream &  file) const
virtual

This saves the trained GaussianMixtureModels model to a file. This overrides the saveModelToFile function in the base class.

Parameters
fstream&file: a reference to the file the GaussianMixtureModels model will be saved to
Returns
returns true if the model was saved successfully, false otherwise

Definition at line 313 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::setNumRestarts ( const UINT  numRestarts)

The Gaussian Mixture Model learning algorithm can fail if the algorithm starts at a bad initial starting point (which is picked at random). To improve the robustness of the learning algorithm, the learning algorithm can be restarted from a new random location if it fails on the previous attempt. The numRestarts parameter controls the maximum number of times the learning algorithm can reattempt to converge, a value of 0 indicates that the learning algorithm must converge on the first attempt, otherwise training will fail.

Returns
returns true if the parameter was updated successfully, false otherwise

Definition at line 546 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::train_ ( MatrixFloat &  trainingData)
virtual

This is the main training interface for referenced MatrixFloat data. It overrides the train_ function in the ML base class.

Parameters
trainingDataa reference to the training data that will be used to train the ML model
Returns
returns true if the model was successfully trained, false otherwise

Reimplemented from Clusterer.

Definition at line 123 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::train_ ( ClassificationData &  trainingData)
virtual

This is the main training interface for reference ClassificationData data. It overrides the train_ function in the ML base class.

Parameters
trainingDataa reference to the training data that will be used to train the ML model
Returns
returns true if the model was successfully trained, false otherwise

Reimplemented from Clusterer.

Definition at line 250 of file GaussianMixtureModels.cpp.

bool GaussianMixtureModels::train_ ( UnlabelledData &  trainingData)
virtual

This is the main training interface for reference UnlabelledData data. It overrides the train_ function in the ML base class.

Parameters
trainingDataa reference to the training data that will be used to train the ML model
Returns
returns true if the model was successfully trained, false otherwise

Reimplemented from Clusterer.

Definition at line 255 of file GaussianMixtureModels.cpp.

The documentation for this class was generated from the following files: