MLBase Class Reference

This is the main base class that all GRT machine learning algorithms should inherit from. More...

#include <MLBase.h>

Inheritance diagram for MLBase:

Public Types
enum	BaseType {   BASE_TYPE_NOT_SET =0, CLASSIFIER, REGRESSIFIER, CLUSTERER,   PRE_PROCSSING, POST_PROCESSING, FEATURE_EXTRACTION, CONTEXT }

Public Member Functions
	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_ (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_ (UnlabelledData &trainingData)
virtual bool	train (MatrixFloat data)
virtual bool	train_ (MatrixFloat &data)
virtual bool	predict (VectorFloat inputVector)
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	reset ()
virtual bool	clear ()
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)

Protected Member Functions
bool	saveBaseSettingsToFile (std::fstream &file) const
bool	loadBaseSettingsFromFile (std::fstream &file)

Protected Attributes
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
Static Public Member Functions inherited from GRTBase
static std::string	getGRTVersion (bool returnRevision=true)
static std::string	getGRTRevison ()

Detailed Description

This is the main base class that all GRT machine learning algorithms should inherit from.

A large number of the functions in this class are virtual and simply return false as these functions must be overwridden by the inheriting class.

Definition at line 72 of file MLBase.h.

Constructor & Destructor Documentation

GRT_BEGIN_NAMESPACE MLBase::MLBase	(	const std::string &	id = "",
		const BaseType	type = BASE_TYPE_NOT_SET
	)

Default MLBase Constructor

Parameters

id	the id of the inheriting class
type	the type of the inheriting class (e.g., classifier, regressifier, etc.)

Definition at line 26 of file MLBase.cpp.

MLBase::~MLBase ( void  )

virtual

Default MLBase Destructor

Definition at line 58 of file MLBase.cpp.

Member Function Documentation

bool MLBase::clear ( )

virtual

This is the main clear interface for all the GRT machine learning algorithms. It will completely clear the ML 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 in GestureRecognitionPipeline, SelfOrganizingMap, DTW, HierarchicalClustering, HMM, FFT, FiniteStateMachine, AdaBoost, ANBC, KNN, SVM, RandomForests, DecisionTree, BAG, ClusterTreeNode, FIRFilter, RegressionTreeNode, RBMQuantizer, GMM, ParticleClassifier, KMeansQuantizer, SOMQuantizer, MLP, ClusterTree, Softmax, SwipeDetector, RegressionTree, MinDist, Clusterer, BernoulliRBM, DecisionTreeNode, Node, KMeans, FeatureExtraction, PreProcessing, GaussianMixtureModels, Classifier, Regressifier, DecisionTreeClusterNode, DecisionTreeThresholdNode, ContinuousHiddenMarkovModel, DecisionTreeTripleFeatureNode, Tree, LinearLeastSquares, and MovementDetector.

Definition at line 149 of file MLBase.cpp.

bool MLBase::copyMLBaseVariables

(

const MLBase *

mlBase

)

This copies all the MLBase variables from the instance mlBaseA to the instance mlBaseA.

Parameters

mlBase

a pointer to a MLBase class from which the values will be copied to the instance that calls the function

Returns

returns true if the copy was successfull, false otherwise

Definition at line 62 of file MLBase.cpp.

bool MLBase::enableScaling

(

const bool

useScaling

)

Sets if scaling should be used during the training and prediction phases.

Returns

returns true the scaling parameter was updated, false otherwise

Examples:

ClassificationModulesExamples/ANBCExample/ANBCExample.cpp, ClassificationModulesExamples/KNNExample/KNNExample.cpp, ClassificationModulesExamples/SVMExample/SVMExample.cpp, and RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 308 of file MLBase.cpp.

UINT MLBase::getBatchSize

(

)

const

Gets the batch size. This value controls the number of samples that can be used by the training algorithm.

Returns

returns the batch size

Definition at line 250 of file MLBase.cpp.

bool MLBase::getConverged

(

)

const

Returns true if the training algorithm converged during the most recent training process. This function will return false if the model has not been trained.

Returns

returns true if the training algorithm converged succesfully, false otherwise

Definition at line 298 of file MLBase.cpp.

DataType MLBase::getInputType

(

)

const

Gets the expected input data type for the module

Returns

returns the expected input data type

Definition at line 223 of file MLBase.cpp.

bool MLBase::getIsBaseTypeClassifier

(

)

const

Gets if the derived class type is CLASSIFIER.

Returns

returns true if the derived class type is CLASSIFIER, false otherwise

Definition at line 302 of file MLBase.cpp.

bool MLBase::getIsBaseTypeClusterer

(

)

const

Gets if the derived class type is CLUSTERER.

Returns

returns true if the derived class type is CLUSTERER, false otherwise

Definition at line 306 of file MLBase.cpp.

bool MLBase::getIsBaseTypeRegressifier

(

)

const

Gets if the derived class type is REGRESSIFIER.

Returns

returns true if the derived class type is REGRESSIFIER, false otherwise

Definition at line 304 of file MLBase.cpp.

Float MLBase::getLearningRate

(

)

const

Gets the current learningRate value, this is value used to update the weights at each step of a learning algorithm such as stochastic gradient descent.

Returns

returns the current learningRate value

Definition at line 262 of file MLBase.cpp.

UINT MLBase::getMaxNumEpochs

(

)

const

Gets the maximum number of epochs. This value controls the maximum number of epochs that can be used by the training algorithm. An epoch is a complete iteration of all training samples.

Returns

returns the maximum number of epochs

Definition at line 246 of file MLBase.cpp.

Float MLBase::getMinChange

(

)

const

Gets the minimum change value that controls when the training algorithm should stop.

Returns

returns the minimum change value

UINT MLBase::getMinNumEpochs

(

)

const

Gets the minimum number of epochs. This is the minimum number of epochs that can elapse with no change between two training epochs. An epoch is a complete iteration of all training samples.

Returns

returns the minimum number of epochs

Definition at line 242 of file MLBase.cpp.

MLBase * MLBase::getMLBasePointer

(

)

This functions returns a pointer to the current instance.

Returns

returns a MLBase pointer to the current instance.

Definition at line 423 of file MLBase.cpp.

const MLBase * MLBase::getMLBasePointer

(

)

const

This functions returns a const pointer to the current instance.

Returns

returns a const MLBase pointer to the current instance.

Definition at line 427 of file MLBase.cpp.

bool MLBase::getModel ( std::ostream &  stream ) const

virtual

This function adds the current model to the formatted stream. This function should be overwritten by the derived class.

Parameters

file	a reference to the stream the model will be added to

Returns

returns true if the model was added successfully, false otherwise

Reimplemented in DecisionTree, Node, DecisionTreeClusterNode, DecisionTreeNode, DecisionTreeThresholdNode, DecisionTreeTripleFeatureNode, and Tree.

Definition at line 213 of file MLBase.cpp.

std::string MLBase::getModelAsString ( ) const

virtual

Gets the current model and settings as a std::string.

Returns

returns a std::string containing the model

Reimplemented in GestureRecognitionPipeline.

Definition at line 215 of file MLBase.cpp.

UINT MLBase::getNumInputDimensions

(

)

const

Gets the number of input dimensions in trained model.

Returns

returns the number of input dimensions

Definition at line 231 of file MLBase.cpp.

UINT MLBase::getNumInputFeatures

(

)

const

Gets the number of input dimensions in trained model. This function is now depriciated and will be removed in the future, you should use getNumInputDimensions instead.

Returns

returns the number of input dimensions in the trained model, a value of 0 will be returned if the model has not been trained

Definition at line 229 of file MLBase.cpp.

UINT MLBase::getNumOutputDimensions

(

)

const

Gets the number of output dimensions in trained model.

Returns

returns the number of output dimensions

Definition at line 233 of file MLBase.cpp.

UINT MLBase::getNumRestarts

(

)

const

Gets the number of times a learning algorithm can restart during training.

Returns

returns the number of times a learning algorithm can restart during training

Definition at line 254 of file MLBase.cpp.

UINT MLBase::getNumTrainingIterationsToConverge

(

)

const

Gets the number of training iterations that were required for the algorithm to converge.

Returns

returns the number of training iterations required for the training algorithm to converge, a value of 0 will be returned if the model has not been trained

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 235 of file MLBase.cpp.

DataType MLBase::getOutputType

(

)

const

Gets the expected output data type for the module

Returns

returns the expected output data type

Definition at line 225 of file MLBase.cpp.

bool MLBase::getRandomiseTrainingOrder

(

)

const

Returns true if the order of the training dataset should be randomized at each epoch of training. Randomizing the order of the training dataset stops a learning algorithm from focusing too much on the first few examples in the dataset.

Returns

returns true if the order of the training dataset should be randomized, false otherwise

Float MLBase::getRMSTrainingError

(

)

const

Gets the root mean squared error on the training data during the training phase.

Returns

returns the RMS error (on the training data during the training phase)

Definition at line 266 of file MLBase.cpp.

Float MLBase::getRMSValidationError

(

)

const

Gets the root mean squared error on the validation data during the training phase, this will be zero if no validation set was used.

Returns

returns the RMS error (on the validation data during the training phase)

Definition at line 278 of file MLBase.cpp.

bool MLBase::getScalingEnabled

(

)

const

Gets if the scaling has been enabled.

Returns

returns true if scaling is enabled, false otherwise

Definition at line 300 of file MLBase.cpp.

bool MLBase::getTestingLoggingEnabled

(

)

const

Gets the logging state for the testing log, note this returns the logging state for this specific instance of the training log, not the global TestingLog state.

Returns

returns true if logging is enabled, false otherwise

Definition at line 316 of file MLBase.cpp.

Float MLBase::getTotalSquaredTrainingError

(

)

const

Gets the total squared error on the training data during the training phase.

Returns

returns the total squared error (on the training data during the training phase)

Definition at line 274 of file MLBase.cpp.

bool MLBase::getTrained

(

)

const

Gets if the model for the derived class has been succesfully trained.

Returns

returns true if the model for the derived class has been succesfully trained, false otherwise

Definition at line 294 of file MLBase.cpp.

bool MLBase::getTrainingLoggingEnabled

(

)

const

Gets the logging state for the training log, note this returns the logging state for this specific instance of the training log, not the global TrainingLog state.

Returns

returns true if logging is enabled, false otherwise

Definition at line 312 of file MLBase.cpp.

Vector< TrainingResult > MLBase::getTrainingResults

(

)

const

Gets the training results from the last training phase. Each element in the vector represents the training results from 1 training iteration.

Returns

returns a vector of TrainingResult instances containing the training results from the most recent training phase

Definition at line 431 of file MLBase.cpp.

MLBase::BaseType MLBase::getType

(

)

const

Gets the current ML base type.

Returns

returns an enum representing the current ML base type, this will be one of the BaseType enumerations

Definition at line 227 of file MLBase.cpp.

bool MLBase::getUseValidationSet

(

)

const

Returns true if a validation set should be used for training. If true, then the training dataset will be partitioned into a smaller training dataset and a validation set.

The size of the partition is controlled by the validationSetSize parameter, for example, if the validationSetSize parameter is 20 then 20% of the training data will be used for a validation set leaving 80% of the original data to train the model.

Returns

returns true if a validation set should be used for training, false otherwise

Definition at line 310 of file MLBase.cpp.

Float MLBase::getValidationSetAccuracy

(

)

const

Gets the accuracy of the validation set on the trained model, only valid if the model was trained with useValidationSet=true.

Returns

returns the accuracy of validation set on the trained model

Definition at line 282 of file MLBase.cpp.

VectorFloat MLBase::getValidationSetPrecision

(

)

const

Gets the precision of the validation set on the trained model, only valid if the model was trained with useValidationSet=true.

Returns

returns the precision of the validation set on the trained model

Definition at line 286 of file MLBase.cpp.

VectorFloat MLBase::getValidationSetRecall

(

)

const

Gets the recall of the validation set on the trained model, only valid if the model was trained with useValidationSet=true.

Returns

returns the recall of the validation set on the trained model

Definition at line 290 of file MLBase.cpp.

UINT MLBase::getValidationSetSize

(

)

const

Gets the size (as a percentage) of the validation set (if one should be used). If this value returned 20 this would mean that 20% of the training data would be set aside to create a validation set and the other 80% would be used to actually train the regression model. This will only happen if the useValidationSet parameter is set to true, otherwise 100% of the training data will be used to train the regression model.

Returns

returns the size of the validation set

Definition at line 258 of file MLBase.cpp.

MLBase::GRT_DEPRECATED_MSG	(	"saveModelToFile(std::string filename) is	deprecated,
		use save(const std::string &filename) instead"	,
		virtual bool saveModelToFile(const std::string &filename)	const
	)

Deprecated:

use save(std::string filename) instead

Parameters

the	name of the file to save the model to

Returns

returns true if the model was saved successfully, false otherwise

MLBase::GRT_DEPRECATED_MSG	(	"saveModelToFile(std::fstream &file) is	deprecated,
		use save(std::fstream &file) instead"	,
		virtual bool saveModelToFile(std::fstream &file)	const
	)

Deprecated:

use save(std::fstream &file) instead

Parameters

file	a reference to the file the model will be saved to

Returns

returns true if the model was saved successfully, false otherwise

MLBase::GRT_DEPRECATED_MSG	(	"loadModelFromFile(std::string filename) is	deprecated,
		use load(const std::string &filename) instead"	,
		virtual bool	loadModelFromFileconst std::string &filename
	)

Deprecated:

use load(std::string filename) instead

Parameters

filename

the name of the file to load the model from

Returns

returns true if the model was loaded successfully, false otherwise

MLBase::GRT_DEPRECATED_MSG	(	"loadModelFromFile(std::fstream &file) is	deprecated,
		use load(std::fstream &file) instead"	,
		virtual bool	loadModelFromFilestd::fstream &file
	)

Deprecated:

use load(std::fstream &file) instead

Parameters

file	a reference to the file the model will be loaded from

Returns

returns true if the model was loaded successfully, false otherwise

MLBase::GRT_DEPRECATED_MSG	(	"getRootMeanSquaredTrainingError() is	deprecated,
		use getRMSTrainingError() instead"	,
		Float getRootMeanSquaredTrainingError()	const
	)

Deprecated:

use getRMSTrainingError() instead Gets the root mean squared error on the training data during the training phase.

Returns

returns the RMS error (on the training data during the training phase)

MLBase::GRT_DEPRECATED_MSG	(	"getModelTrained() is	deprecated,
		use getTrained() instead"	,
		bool getModelTrained()	const
	)

Deprecated:

getTrained() function instead This function is now depreciated. You should use the getTrained() function instead.

Returns

returns true if the model for the derived class has been succesfully trained, false otherwise

bool MLBase::load ( const std::string &  filename )

virtual

This saves the model to a file.

Parameters

filename

the name of the file to save the model to

Returns

returns true if the model was saved successfully, false otherwise

Reimplemented in GestureRecognitionPipeline.

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 190 of file MLBase.cpp.

bool MLBase::load ( std::fstream &  file )

virtual

This loads a trained model from a file. This function should be overwritten by the derived class.

Parameters

file	a reference to the file the model will be loaded from

Returns

returns true if the model was loaded successfully, false otherwise

Reimplemented in SelfOrganizingMap, DTW, ClassLabelTimeoutFilter, FFT, HMM, FiniteStateMachine, RandomForests, AdaBoost, DecisionTree, ANBC, KNN, SVM, MLP, SwipeDetector, RegressionTree, BAG, ZeroCrossingCounter, FIRFilter, KMeansQuantizer, RBMQuantizer, GMM, PrincipalComponentAnalysis, SOMQuantizer, Softmax, MinDist, Node, ClassLabelFilter, MovementTrajectoryFeatures, TimeDomainFeatures, Derivative, LowPassFilter, MultidimensionalRegression, BernoulliRBM, HighPassFilter, FFTFeatures, KMeansFeatures, SavitzkyGolayFilter, MovementIndex, LogisticRegression, ClassLabelChangeFilter, DeadZone, DoubleMovingAverageFilter, EnvelopeExtractor, MedianFilter, MovingAverageFilter, LeakyIntegrator, RMSFilter, WeightedAverageFilter, TimeseriesBuffer, LinearRegression, ParticleClassifier, DiscreteHiddenMarkovModel, ContinuousHiddenMarkovModel, and MovementDetector.

Definition at line 205 of file MLBase.cpp.

bool MLBase::loadBaseSettingsFromFile ( std::fstream &  file )

protected

Loads the core base settings from a file.

Returns

returns true if the base settings were loaded, false otherwise

Definition at line 458 of file MLBase.cpp.

bool MLBase::map ( VectorFloat  inputVector )

virtual

This is the main mapping interface for all the GRT machine learning algorithms. By defaut it will call the map_ function, unless it is overwritten by the derived class.

Parameters

inputVector

the input vector for mapping/regression

Returns

returns true if the mapping was completed succesfully, false otherwise (the base class always returns false)

Definition at line 143 of file MLBase.cpp.

bool MLBase::map_ ( VectorFloat &  inputVector )

virtual

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

Parameters

inputVector

a reference to the input vector for mapping/regression

Returns

returns true if the mapping was completed succesfully, false otherwise (the base class always returns false)

Reimplemented in SelfOrganizingMap.

Definition at line 145 of file MLBase.cpp.

bool MLBase::notifyTestResultsObservers

(

const TestInstanceResult &

data

)

Notifies all observers that have subscribed to the test results observer manager.

Parameters

data	stores the test results data for the current update

Returns

returns true if all the observers were notified, false otherwise

Definition at line 419 of file MLBase.cpp.

bool MLBase::notifyTrainingResultsObservers

(

const TrainingResult &

data

)

Notifies all observers that have subscribed to the training results observer manager.

Parameters

data	stores the training results data for the current update

Returns

returns true if all the observers were notified, false otherwise

Definition at line 415 of file MLBase.cpp.

bool MLBase::predict ( VectorFloat  inputVector )

virtual

This is the main prediction interface for all the GRT machine learning algorithms. By defaut it will call the predict_ function, unless it is overwritten by the derived class.

Parameters

inputVector

the new input vector for prediction

Returns

returns true if the prediction was completed succesfully, false otherwise (the base class always returns false)

Examples:

ClassificationModulesExamples/AdaBoostExample/AdaBoostExample.cpp, ClassificationModulesExamples/ANBCExample/ANBCExample.cpp, ClassificationModulesExamples/BAGExample/BAGExample.cpp, ClassificationModulesExamples/DecisionTreeExample/DecisionTreeExample.cpp, ClassificationModulesExamples/DTWExample/DTWExample.cpp, ClassificationModulesExamples/GMMExample/GMMExample.cpp, ClassificationModulesExamples/MinDistExample/MinDistExample.cpp, ClassificationModulesExamples/RandomForestsExample/RandomForestsExample.cpp, ClassificationModulesExamples/SoftmaxExample/SoftmaxExample.cpp, ClassificationModulesExamples/SVMExample/SVMExample.cpp, RegressionModulesExamples/LinearRegressionExample/LinearRegressionExample.cpp, RegressionModulesExamples/LogisticRegressionExample/LogisticRegressionExample.cpp, RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp, and RegressionModulesExamples/MultidimensionalRegressionExample/MultidimensionalRegressionExample.cpp.

Definition at line 135 of file MLBase.cpp.

bool MLBase::predict ( MatrixFloat  inputMatrix )

virtual

This is the prediction interface for time series data. By defaut it will call the predict_ function, unless it is overwritten by the derived class.

Parameters

inputMatrix

the new input matrix for prediction

Returns

returns true if the prediction was completed succesfully, false otherwise (the base class always returns false)

Definition at line 139 of file MLBase.cpp.

bool MLBase::predict_ ( VectorFloat &  inputVector )

virtual

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

Parameters

inputVector

a 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 in GestureRecognitionPipeline, DTW, KMeans, AdaBoost, SVM, HMM, KNN, GaussianMixtureModels, RandomForests, FiniteStateMachine, DecisionTree, ANBC, MLP, GMM, MultidimensionalRegression, ClusterTree, Softmax, SwipeDetector, RegressionTree, BAG, MinDist, LogisticRegression, LinearRegression, ParticleClassifier, RegressionTreeNode, DecisionTreeClusterNode, DecisionTreeThresholdNode, ClusterTreeNode, DecisionTreeTripleFeatureNode, BernoulliRBM, Node, ContinuousHiddenMarkovModel, and MovementDetector.

Definition at line 137 of file MLBase.cpp.

bool MLBase::predict_ ( MatrixFloat &  inputMatrix )

virtual

This is the prediction interface for time series data. This should be overwritten by the derived class.

Parameters

inputMatrix

a reference to the new input matrix for prediction

Returns

returns true if the prediction was completed succesfully, false otherwise (the base class always returns false)

Reimplemented in GestureRecognitionPipeline, DTW, HMM, and ContinuousHiddenMarkovModel.

Definition at line 141 of file MLBase.cpp.

bool MLBase::print ( ) const

virtual

This is the main print interface for all the GRT machine learning algorithms. This should be overwritten by the derived class. It will print the model and settings to the display log.

Returns

returns true if the model was printed succesfully, false otherwise (the base class always returns true)

Reimplemented in HMM, FiniteStateMachine, ClusterTreeNode, RegressionTreeNode, RandomForests, BernoulliRBM, MLP, ClusterTree, RegressionTree, Node, DiscreteHiddenMarkovModel, ContinuousHiddenMarkovModel, DecisionTreeClusterNode, DecisionTreeThresholdNode, DecisionTreeTripleFeatureNode, and Tree.

Examples:

ClassificationModulesExamples/DecisionTreeExample/DecisionTreeExample.cpp.

Definition at line 165 of file MLBase.cpp.

bool MLBase::registerTestResultsObserver

(

Observer< TestInstanceResult > &

observer

)

Registers the observer with the test result observer manager. The observer will then be notified when any new test result is computed.

Parameters

observer

the observer you want to register with the learning algorithm

Returns

returns true the observer was added, false otherwise

Definition at line 395 of file MLBase.cpp.

bool MLBase::registerTrainingResultsObserver

(

Observer< TrainingResult > &

observer

)

Registers the observer with the training result observer manager. The observer will then be notified when any new training result is computed.

Parameters

observer

the observer you want to register with the learning algorithm

Returns

returns true the observer was added, false otherwise

Definition at line 391 of file MLBase.cpp.

bool MLBase::removeAllTestObservers

(

)

Removes all observers from the training result observer manager.

Returns

returns true if all the observers were removed, false otherwise

Definition at line 411 of file MLBase.cpp.

bool MLBase::removeAllTrainingObservers

(

)

Removes all observers from the training result observer manager.

Returns

returns true if all the observers were removed, false otherwise

Definition at line 407 of file MLBase.cpp.

bool MLBase::removeTestResultsObserver

(

const Observer< TestInstanceResult > &

observer

)

Removes the observer from the test result observer manager.

Parameters

observer

the observer you want to remove from the learning algorithm

Returns

returns true if the observer was removed, false otherwise

Definition at line 403 of file MLBase.cpp.

bool MLBase::removeTrainingResultsObserver

(

const Observer< TrainingResult > &

observer

)

Removes the observer from the training result observer manager.

Parameters

observer

the observer you want to remove from the learning algorithm

Returns

returns true if the observer was removed, false otherwise

Definition at line 399 of file MLBase.cpp.

bool MLBase::reset ( )

virtual

This is the main reset interface for all the GRT machine learning algorithms. It should be used to reset the model (i.e. set all values back to default settings). If you want to completely clear the model (i.e. clear any learned weights or values) then you should use the clear function.

Returns

returns true if the derived class was reset succesfully, false otherwise (the base class always returns true)

Reimplemented in GestureRecognitionPipeline, SelfOrganizingMap, DTW, HierarchicalClustering, ClassLabelTimeoutFilter, FFT, HMM, FiniteStateMachine, ParticleClassifier, SwipeDetector, ANBC, ZeroCrossingCounter, BAG, ClassLabelFilter, RBMQuantizer, FIRFilter, KMeansQuantizer, MovementTrajectoryFeatures, SOMQuantizer, TimeDomainFeatures, Derivative, LowPassFilter, HighPassFilter, FFTFeatures, Clusterer, SavitzkyGolayFilter, KMeansFeatures, MovementIndex, BernoulliRBM, ClassLabelChangeFilter, DeadZone, DoubleMovingAverageFilter, MedianFilter, MovingAverageFilter, EnvelopeExtractor, LeakyIntegrator, RMSFilter, WeightedAverageFilter, TimeseriesBuffer, Context, KMeans, PreProcessing, GaussianMixtureModels, Classifier, Regressifier, DiscreteHiddenMarkovModel, Gate, ContinuousHiddenMarkovModel, and MovementDetector.

Definition at line 147 of file MLBase.cpp.

bool MLBase::save ( const std::string &  filename ) const

virtual

This saves the model to a file.

Parameters

filename

the name of the file to save the model to

Returns

returns true if the model was saved successfully, false otherwise

Reimplemented in GestureRecognitionPipeline.

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp.

Definition at line 167 of file MLBase.cpp.

bool MLBase::save ( std::fstream &  file ) const

virtual

This saves the trained model to a file. This function should be overwritten by the derived class.

Parameters

file	a reference to the file the model will be saved to

Returns

returns true if the model was saved successfully, false otherwise

Reimplemented in SelfOrganizingMap, DTW, ClassLabelTimeoutFilter, FFT, HMM, FiniteStateMachine, RandomForests, AdaBoost, DecisionTree, ANBC, KNN, SVM, MLP, SwipeDetector, RegressionTree, BAG, ZeroCrossingCounter, FIRFilter, PrincipalComponentAnalysis, KMeansQuantizer, RBMQuantizer, GMM, SOMQuantizer, Softmax, Node, MinDist, ClassLabelFilter, MovementTrajectoryFeatures, TimeDomainFeatures, BernoulliRBM, Derivative, LowPassFilter, MultidimensionalRegression, HighPassFilter, FFTFeatures, KMeansFeatures, SavitzkyGolayFilter, MovementIndex, LogisticRegression, ClassLabelChangeFilter, DeadZone, DoubleMovingAverageFilter, EnvelopeExtractor, MedianFilter, MovingAverageFilter, LeakyIntegrator, RMSFilter, WeightedAverageFilter, TimeseriesBuffer, LinearRegression, ParticleClassifier, DiscreteHiddenMarkovModel, ContinuousHiddenMarkovModel, and MovementDetector.

Definition at line 182 of file MLBase.cpp.

bool MLBase::saveBaseSettingsToFile ( std::fstream &  file ) const

protected

Saves the core base settings to a file.

Returns

returns true if the base settings were saved, false otherwise

Definition at line 435 of file MLBase.cpp.

bool MLBase::setBatchSize

(

const UINT

batchSize

)

Sets the batch size used during the training phase.

Parameters

batchSize

the batch size

Returns

returns true if the value was updated successfully, false otherwise

Definition at line 334 of file MLBase.cpp.

bool MLBase::setLearningRate

(

const Float

learningRate

)

Sets the learningRate. This is used to update the weights at each step of learning algorithms such as stochastic gradient descent. The learningRate value must be greater than zero.

Parameters

learningRate

the learningRate value used during the training phase, must be greater than zero

Returns

returns true if the value was updated successfully, false otherwise

Examples:

RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 353 of file MLBase.cpp.

bool MLBase::setMaxNumEpochs

(

const UINT

maxNumEpochs

)

Sets the maximum number of epochs (a complete iteration of all training samples) that can be run during the training phase. The maxNumIterations value must be greater than zero.

Parameters

maxNumIterations

the maximum number of iterations value, must be greater than zero

Returns

returns true if the value was updated successfully, false otherwise

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp, ClusteringModulesExamples/KMeansExample/KMeansExample.cpp, and RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 320 of file MLBase.cpp.

bool MLBase::setMinChange

(

const Float

minChange

)

Sets the minimum change that must be achieved between two training epochs for the training to continue. The minChange value must be greater than zero.

Parameters

minChange

the minimum change value, must be greater than zero

Returns

returns true if the value was updated successfully, false otherwise

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp, ClusteringModulesExamples/KMeansExample/KMeansExample.cpp, and RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 344 of file MLBase.cpp.

bool MLBase::setMinNumEpochs

(

const UINT

minNumEpochs

)

Sets the minimum number of epochs (a complete iteration of all training samples) that can elapse with no change between two training epochs.

Parameters

minNumEpochs

the minimum number of epochs that can elapse with no change between two training epochs

Returns

returns true if the value was updated successfully, false otherwise

Examples:

ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp, and ClusteringModulesExamples/KMeansExample/KMeansExample.cpp.

Definition at line 329 of file MLBase.cpp.

bool MLBase::setNumRestarts

(

const UINT

numRestarts

)

Sets the number of times a learning algorithm can restart during training. This is used to restart the training for algorithms that can get stuck if they start with bad random values.

Parameters

numRestarts

number of times a learning algorithm can restart during training

Returns

returns true if the value was updated successfully, false otherwise

Examples:

RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 339 of file MLBase.cpp.

bool MLBase::setRandomiseTrainingOrder

(

const bool

randomiseTrainingOrder

)

Sets if the order of the training dataset should be randomized at each epoch of training. Randomizing the order of the training dataset stops a learning algorithm from focusing too much on the first few examples in the dataset.

Parameters

randomiseTrainingOrder

if true then the order in which training samples are supplied to a learning algorithm will be randomised

Returns

returns true if the parameter was updated, false otherwise

Examples:

RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 378 of file MLBase.cpp.

bool MLBase::setTestingLoggingEnabled

(

const bool

loggingEnabled

)

Sets if testing logging is enabled/disabled for this specific ML instance. If you want to enable/disable testing logging globally, then you should use the TestingLog::enableLogging( bool ) function.

Parameters

loggingEnabled

if true then training logging will be enabled, if false then training logging will be disabled

Returns

returns true if the parameter was updated, false otherwise

Definition at line 387 of file MLBase.cpp.

bool MLBase::setTrainingLoggingEnabled

(

const bool

loggingEnabled

)

Sets if training logging is enabled/disabled for this specific ML instance. If you want to enable/disable training logging globally, then you should use the TrainingLog::enableLogging( bool ) function.

Parameters

loggingEnabled

if true then training logging will be enabled, if false then training logging will be disabled

Returns

returns true if the parameter was updated, false otherwise

Definition at line 383 of file MLBase.cpp.

bool MLBase::setUseValidationSet

(

const bool

useValidationSet

)

Sets the size of the validation set used by some learning algorithms for training. This value represents the percentage of the main dataset that will be used for training. For example, if the validationSetSize parameter is 20 then 20% of the training data will be used for a validation set leaving 80% of the original data to train the model.

Parameters

validationSetSize

the new validation set size (as a percentage)

Returns

returns true if the validationSetSize parameter was updated, false otherwise

Examples:

RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 373 of file MLBase.cpp.

bool MLBase::setValidationSetSize

(

const UINT

validationSetSize

)

Sets the size of the validation set used by some learning algorithms for training. This value represents the percentage of the main dataset that will be used for training. For example, if the validationSetSize parameter is 20 then 20% of the training data will be used for a validation set leaving 80% of the original data to train the model.

Parameters

validationSetSize

the new validation set size (as a percentage)

Returns

returns true if the validationSetSize parameter was updated, false otherwise

Examples:

RegressionModulesExamples/MLPRegressionExample/MLPRegressionExample.cpp.

Definition at line 361 of file MLBase.cpp.

bool MLBase::train ( ClassificationData  trainingData )

virtual

This is the main training interface for ClassificationData. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in HMM.

Examples:

ClassificationModulesExamples/AdaBoostExample/AdaBoostExample.cpp, ClassificationModulesExamples/ANBCExample/ANBCExample.cpp, ClassificationModulesExamples/BAGExample/BAGExample.cpp, ClassificationModulesExamples/DecisionTreeExample/DecisionTreeExample.cpp, ClassificationModulesExamples/DTWExample/DTWExample.cpp, ClassificationModulesExamples/GMMExample/GMMExample.cpp, ClassificationModulesExamples/KNNExample/KNNExample.cpp, ClassificationModulesExamples/MinDistExample/MinDistExample.cpp, ClassificationModulesExamples/RandomForestsExample/RandomForestsExample.cpp, ClassificationModulesExamples/SoftmaxExample/SoftmaxExample.cpp, ClassificationModulesExamples/SVMExample/SVMExample.cpp, ClusteringModulesExamples/GaussianMixtureModelsExample/GaussianMixtureModelsExample.cpp, and ClusteringModulesExamples/KMeansExample/KMeansExample.cpp.

Definition at line 107 of file MLBase.cpp.

bool MLBase::train ( RegressionData  trainingData )

virtual

This is the main training interface for regression data. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train a new regression model

Returns

returns true if a new regression model was trained, false otherwise

Definition at line 111 of file MLBase.cpp.

bool MLBase::train ( RegressionData  trainingData, RegressionData  validationData  )

virtual

This is the main training interface for training a regression model using a training and validation dataset. This should be overwritten by the derived class. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData	the training data that will be used to train a new regression model
validationData	the validation data that will be used to validate the regression model

Returns

returns true if a new regression model was trained, false otherwise

Definition at line 115 of file MLBase.cpp.

bool MLBase::train ( TimeSeriesClassificationData  trainingData )

virtual

This is the main training interface for TimeSeriesClassificationData. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Definition at line 119 of file MLBase.cpp.

bool MLBase::train ( ClassificationDataStream  trainingData )

virtual

This is the main training interface for ClassificationDataStream. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Definition at line 123 of file MLBase.cpp.

bool MLBase::train ( UnlabelledData  trainingData )

virtual

This is the main training interface for UnlabelledData. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Definition at line 127 of file MLBase.cpp.

bool MLBase::train ( MatrixFloat  data )

virtual

This is the main training interface for MatrixFloat data. By default it will call the train_ function, unless it is overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Definition at line 131 of file MLBase.cpp.

bool MLBase::train_ ( ClassificationData &  trainingData )

virtual

This is the main training interface for referenced ClassificationData. This should be overwritten by the derived class.

Parameters

trainingData

a reference to the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in SelfOrganizingMap, HierarchicalClustering, KMeansQuantizer, RBMQuantizer, SOMQuantizer, GestureRecognitionPipeline, KMeansFeatures, KMeans, AdaBoost, SVM, KNN, RandomForests, GaussianMixtureModels, DecisionTree, ANBC, GMM, Softmax, SwipeDetector, BAG, MinDist, MLP, Clusterer, and FiniteStateMachine.

Definition at line 109 of file MLBase.cpp.

bool MLBase::train_ ( RegressionData &  trainingData )

virtual

This is the main training interface for all the regression algorithms. This should be overwritten by the derived class.

Parameters

trainingData

the training data that will be used to train a new regression model

Returns

returns true if a new regression model was trained, false otherwise

Reimplemented in GestureRecognitionPipeline, MLP, MultidimensionalRegression, RegressionTree, LogisticRegression, and LinearRegression.

Definition at line 113 of file MLBase.cpp.

bool MLBase::train_ ( RegressionData &  trainingData, RegressionData &  validationData  )

virtual

This is the main training interface for training a regression model using a training and validation dataset. This should be overwritten by the derived class.

Parameters

trainingData	the training data that will be used to train a new regression model
validationData	the validation data that will be used to validate the regression model

Returns

returns true if a new regression model was trained, false otherwise

Reimplemented in GestureRecognitionPipeline.

Definition at line 117 of file MLBase.cpp.

bool MLBase::train_ ( TimeSeriesClassificationData &  trainingData )

virtual

This is the main training interface for referenced TimeSeriesClassificationData. This should be overwritten by the derived class.

Parameters

trainingData

a reference to the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in GestureRecognitionPipeline, DTW, KMeansQuantizer, RBMQuantizer, SOMQuantizer, KMeansFeatures, HMM, FiniteStateMachine, and ParticleClassifier.

Definition at line 121 of file MLBase.cpp.

bool MLBase::train_ ( ClassificationDataStream &  trainingData )

virtual

This is the main training interface for referenced ClassificationDataStream. This should be overwritten by the derived class.

Parameters

trainingData

a reference to the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in GestureRecognitionPipeline, KMeansQuantizer, RBMQuantizer, SOMQuantizer, KMeansFeatures, and FiniteStateMachine.

Definition at line 125 of file MLBase.cpp.

bool MLBase::train_ ( UnlabelledData &  trainingData )

virtual

This is the main training interface for referenced UnlabelledData. This should be overwritten by the derived class.

Parameters

trainingData

a reference to the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in SelfOrganizingMap, GestureRecognitionPipeline, HierarchicalClustering, KMeansQuantizer, RBMQuantizer, SOMQuantizer, KMeansFeatures, KMeans, GaussianMixtureModels, and Clusterer.

Definition at line 129 of file MLBase.cpp.

bool MLBase::train_ ( MatrixFloat &  data )

virtual

This is the main training interface for referenced MatrixFloat data. This should be overwritten by the derived class.

Parameters

trainingData

a reference to the training data that will be used to train the ML model

Returns

returns true if the classifier was successfully trained, false otherwise

Reimplemented in SelfOrganizingMap, HierarchicalClustering, KMeansQuantizer, RBMQuantizer, SOMQuantizer, KMeansFeatures, KMeans, GaussianMixtureModels, ClusterTree, BernoulliRBM, and Clusterer.

Definition at line 133 of file MLBase.cpp.

---

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

• MLBase.h
• MLBase.cpp