GestureRecognitionToolkit  Version: 1.0 Revision: 04-03-15
The Gesture Recognition Toolkit (GRT) is a cross-platform, open-source, c++ machine learning library for real-time gesture recognition.
GRT::ClusterTree Class Reference
Inheritance diagram for GRT::ClusterTree:
GRT::Tree GRT::Clusterer GRT::GRTBase GRT::MLBase GRT::GRTBase GRT::Observer< TrainingResult > GRT::Observer< TestInstanceResult >

Public Member Functions

 ClusterTree (const UINT numSplittingSteps=100, const UINT minNumSamplesPerNode=5, const UINT maxDepth=10, const bool removeFeaturesAtEachSpilt=false, const UINT trainingMode=BEST_ITERATIVE_SPILT, const bool useScaling=false, const double minRMSErrorPerNode=0.01)
 
 ClusterTree (const ClusterTree &rhs)
 
virtual ~ClusterTree (void)
 
ClusterTreeoperator= (const ClusterTree &rhs)
 
virtual bool deepCopyFrom (const Clusterer *cluster)
 
virtual bool train_ (MatrixDouble &trainingData)
 
virtual bool predict_ (VectorDouble &inputVector)
 
virtual bool clear ()
 
virtual bool print () const
 
virtual bool saveModelToFile (fstream &file) const
 
virtual bool loadModelFromFile (fstream &file)
 
ClusterTreeNodedeepCopyTree () const
 
const ClusterTreeNodegetTree () const
 
UINT getPredictedClusterLabel () const
 
double getMinRMSErrorPerNode () const
 
bool setMinRMSErrorPerNode (const double minRMSErrorPerNode)
 
- Public Member Functions inherited from GRT::Tree
 Tree (const UINT numSplittingSteps=100, const UINT minNumSamplesPerNode=5, const UINT maxDepth=10, const bool removeFeaturesAtEachSpilt=false, const UINT trainingMode=BEST_ITERATIVE_SPILT)
 
virtual ~Tree (void)
 
virtual bool getModel (ostream &stream) const
 
const NodegetTree () const
 
UINT getTrainingMode () const
 
UINT getNumSplittingSteps () const
 
UINT getMinNumSamplesPerNode () const
 
UINT getMaxDepth () const
 
UINT getPredictedNodeID () const
 
bool getRemoveFeaturesAtEachSpilt () const
 
bool setTrainingMode (const UINT trainingMode)
 
bool setNumSplittingSteps (const UINT numSplittingSteps)
 
bool setMinNumSamplesPerNode (const UINT minNumSamplesPerNode)
 
bool setMaxDepth (const UINT maxDepth)
 
bool setRemoveFeaturesAtEachSpilt (const bool removeFeaturesAtEachSpilt)
 
- Public Member Functions inherited from GRT::GRTBase
 GRTBase (void)
 
virtual ~GRTBase (void)
 
bool copyGRTBaseVariables (const GRTBase *GRTBase)
 
string getClassType () const
 
string getLastWarningMessage () const
 
string getLastErrorMessage () const
 
string getLastInfoMessage () const
 
GRTBasegetGRTBasePointer ()
 
const GRTBasegetGRTBasePointer () const
 
- Public Member Functions inherited from GRT::Clusterer
 Clusterer (void)
 
virtual ~Clusterer (void)
 
bool copyBaseVariables (const Clusterer *Clusterer)
 
virtual bool train_ (ClassificationData &trainingData)
 
virtual bool train_ (UnlabelledData &trainingData)
 
virtual bool reset ()
 
bool getConverged () const
 
UINT getNumClusters () const
 
UINT getPredictedClusterLabel () const
 
double getMaximumLikelihood () const
 
double getBestDistance () const
 
VectorDouble getClusterLikelihoods () const
 
VectorDouble getClusterDistances () const
 
vector< UINT > getClusterLabels () const
 
string getClustererType () const
 
bool setNumClusters (const UINT numClusters)
 
ClusterercreateNewInstance () const
 
ClustererdeepCopy () const
 
const ClusterergetBaseClusterer () const
 
- Public Member Functions inherited from GRT::MLBase
 MLBase (void)
 
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 (TimeSeriesClassificationData trainingData)
 
virtual bool train_ (TimeSeriesClassificationData &trainingData)
 
virtual bool train (TimeSeriesClassificationDataStream trainingData)
 
virtual bool train_ (TimeSeriesClassificationDataStream &trainingData)
 
virtual bool train (UnlabelledData trainingData)
 
virtual bool train (MatrixDouble data)
 
virtual bool predict (VectorDouble inputVector)
 
virtual bool predict (MatrixDouble inputMatrix)
 
virtual bool predict_ (MatrixDouble &inputMatrix)
 
virtual bool map (VectorDouble inputVector)
 
virtual bool map_ (VectorDouble &inputVector)
 
virtual bool save (const string filename) const
 
virtual bool load (const string filename)
 
virtual bool saveModelToFile (string filename) const
 
virtual bool loadModelFromFile (string filename)
 
virtual bool getModel (ostream &stream) const
 
double scale (const double &x, const double &minSource, const double &maxSource, const double &minTarget, const double &maxTarget, const bool constrain=false)
 
virtual string getModelAsString () const
 
UINT getBaseType () const
 
UINT getNumInputFeatures () const
 
UINT getNumInputDimensions () const
 
UINT getNumOutputDimensions () const
 
UINT getMinNumEpochs () const
 
UINT getMaxNumEpochs () const
 
UINT getValidationSetSize () const
 
UINT getNumTrainingIterationsToConverge () const
 
double getMinChange () const
 
double getLearningRate () const
 
double getRootMeanSquaredTrainingError () const
 
double getTotalSquaredTrainingError () const
 
bool getUseValidationSet () const
 
bool getRandomiseTrainingOrder () const
 
bool getTrained () const
 
bool getModelTrained () const
 
bool getScalingEnabled () const
 
bool getIsBaseTypeClassifier () const
 
bool getIsBaseTypeRegressifier () const
 
bool getIsBaseTypeClusterer () const
 
bool enableScaling (bool useScaling)
 
bool setMaxNumEpochs (const UINT maxNumEpochs)
 
bool setMinNumEpochs (const UINT minNumEpochs)
 
bool setMinChange (const double minChange)
 
bool setLearningRate (double learningRate)
 
bool setUseValidationSet (const bool useValidationSet)
 
bool setValidationSetSize (const UINT validationSetSize)
 
bool setRandomiseTrainingOrder (const bool randomiseTrainingOrder)
 
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)
 
MLBasegetMLBasePointer ()
 
const MLBasegetMLBasePointer () const
 
vector< TrainingResultgetTrainingResults () const
 
- Public Member Functions inherited from GRT::Observer< TrainingResult >
virtual void notify (const TrainingResult &data)
 
- Public Member Functions inherited from GRT::Observer< TestInstanceResult >
virtual void notify (const TestInstanceResult &data)
 

Protected Member Functions

ClusterTreeNodebuildTree (const MatrixDouble &trainingData, ClusterTreeNode *parent, vector< UINT > features, UINT &clusterLabel, UINT nodeID)
 
bool computeBestSpilt (const MatrixDouble &trainingData, const vector< UINT > &features, UINT &featureIndex, double &threshold, double &minError)
 
bool computeBestSpiltBestIterativeSpilt (const MatrixDouble &trainingData, const vector< UINT > &features, UINT &featureIndex, double &threshold, double &minError)
 
bool computeBestSpiltBestRandomSpilt (const MatrixDouble &trainingData, const vector< UINT > &features, UINT &featureIndex, double &threshold, double &minError)
 
- Protected Member Functions inherited from GRT::GRTBase
double SQR (const double &x) const
 
- Protected Member Functions inherited from GRT::Clusterer
bool saveClustererSettingsToFile (fstream &file) const
 
bool loadClustererSettingsFromFile (fstream &file)
 
- Protected Member Functions inherited from GRT::MLBase
bool saveBaseSettingsToFile (fstream &file) const
 
bool loadBaseSettingsFromFile (fstream &file)
 

Protected Attributes

double minRMSErrorPerNode
 
- Protected Attributes inherited from GRT::Tree
UINT trainingMode
 
UINT numSplittingSteps
 
UINT minNumSamplesPerNode
 
UINT maxDepth
 
bool removeFeaturesAtEachSpilt
 
Nodetree
 
- Protected Attributes inherited from GRT::GRTBase
string classType
 
DebugLog debugLog
 
ErrorLog errorLog
 
InfoLog infoLog
 
TrainingLog trainingLog
 
TestingLog testingLog
 
WarningLog warningLog
 
- Protected Attributes inherited from GRT::Clusterer
string clustererType
 
UINT numClusters
 Number of clusters in the model.
 
UINT predictedClusterLabel
 Stores the predicted cluster label from the most recent predict( )
 
double maxLikelihood
 
double bestDistance
 
VectorDouble clusterLikelihoods
 
VectorDouble clusterDistances
 
vector< UINT > clusterLabels
 
bool converged
 
vector< MinMaxranges
 
- Protected Attributes inherited from GRT::MLBase
bool trained
 
bool useScaling
 
UINT baseType
 
UINT numInputDimensions
 
UINT numOutputDimensions
 
UINT numTrainingIterationsToConverge
 
UINT minNumEpochs
 
UINT maxNumEpochs
 
UINT validationSetSize
 
double learningRate
 
double minChange
 
double rootMeanSquaredTrainingError
 
double totalSquaredTrainingError
 
bool useValidationSet
 
bool randomiseTrainingOrder
 
Random random
 
vector< TrainingResulttrainingResults
 
TrainingResultsObserverManager trainingResultsObserverManager
 
TestResultsObserverManager testResultsObserverManager
 

Static Protected Attributes

static RegisterClustererModule< ClusterTreeregisterModule
 

Additional Inherited Members

- Public Types inherited from GRT::Tree
enum  TrainingMode { BEST_ITERATIVE_SPILT =0, BEST_RANDOM_SPLIT, NUM_TRAINING_MODES }
 
- Public Types inherited from GRT::Clusterer
typedef std::map< string, Clusterer *(*)() > StringClustererMap
 
- Public Types inherited from GRT::MLBase
enum  BaseTypes { BASE_TYPE_NOT_SET =0, CLASSIFIER, REGRESSIFIER, CLUSTERER }
 
- Static Public Member Functions inherited from GRT::GRTBase
static string getGRTVersion (bool returnRevision=true)
 
static string getGRTRevison ()
 
- Static Public Member Functions inherited from GRT::Clusterer
static ClusterercreateInstanceFromString (string const &ClustererType)
 
static vector< string > getRegisteredClusterers ()
 
- Static Protected Member Functions inherited from GRT::Clusterer
static StringClustererMapgetMap ()
 

Detailed Description

Definition at line 40 of file ClusterTree.h.

Constructor & Destructor Documentation

GRT::ClusterTree::ClusterTree ( const UINT  numSplittingSteps = 100,
const UINT  minNumSamplesPerNode = 5,
const UINT  maxDepth = 10,
const bool  removeFeaturesAtEachSpilt = false,
const UINT  trainingMode = BEST_ITERATIVE_SPILT,
const bool  useScaling = false,
const double  minRMSErrorPerNode = 0.01 
)

Default Constructor

Parameters
UINTnumSplittingSteps: sets the number of steps that will be used to search for the best spliting value for each node. Default value = 100
UINTminNumSamplesPerNode: sets the minimum number of samples that are allowed per node, if the number of samples is below that, the node will become a leafNode. Default value = 5
UINTmaxDepth: sets the maximum depth of the tree. Default value = 10
boolremoveFeaturesAtEachSpilt: sets if a feature is removed at each spilt so it can not be used again. Default value = false
UINTtrainingMode: sets the training mode, this should be one of the TrainingMode enums. Default value = BEST_ITERATIVE_SPILT
booluseScaling: sets if the training and real-time data should be scaled between [0 1]. Default value = false
constdouble minRMSErrorPerNode: sets the minimum RMS error that allowed per node, if the RMS error is below that, the node will become a leafNode. Default value = 0.01

Definition at line 31 of file ClusterTree.cpp.

GRT::ClusterTree::ClusterTree ( const ClusterTree rhs)

Defines the copy constructor.

Parameters
constClusterTree &rhs: the instance from which all the data will be copied into this instance

Definition at line 48 of file ClusterTree.cpp.

GRT::ClusterTree::~ClusterTree ( void  )
virtual

Default Destructor

Definition at line 60 of file ClusterTree.cpp.

Member Function Documentation

bool GRT::ClusterTree::clear ( )
virtual

This overrides the clear function in the Regressifier base class. 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 module was cleared succesfully, false otherwise

Reimplemented from GRT::Tree.

Definition at line 207 of file ClusterTree.cpp.

bool GRT::ClusterTree::deepCopyFrom ( const Clusterer cluster)
virtual

This is required for the Gesture Recognition Pipeline for when the pipeline.setRegressifier(...) method is called. It clones the data from the Base Class Regressifier pointer (which should be pointing to an RegressionTree instance) into this instance

Parameters
Regressifier*regressifier: a pointer to the Regressifier Base Class, this should be pointing to another RegressionTree instance
Returns
returns true if the clone was successfull, false otherwise

Reimplemented from GRT::Clusterer.

Definition at line 89 of file ClusterTree.cpp.

ClusterTreeNode * GRT::ClusterTree::deepCopyTree ( ) const
virtual

Deep copies the regression tree, returning a pointer to the new regression tree. The user is in charge of cleaning up the memory so must delete the pointer when they no longer need it. NULL will be returned if the tree could not be copied.

Returns
returns a pointer to a deep copy of the regression tree

Reimplemented from GRT::Tree.

Definition at line 372 of file ClusterTree.cpp.

double GRT::ClusterTree::getMinRMSErrorPerNode ( ) const

Gets the minimum root mean squared error value that needs to be exceeded for the tree to continue growing at a specific node. If the RMS error is below this value then the node will be made into a leaf node.

Returns
returns the minimum RMS error per node

Definition at line 389 of file ClusterTree.cpp.

UINT GRT::ClusterTree::getPredictedClusterLabel ( ) const

Gets the predicted cluster label from the most recent call to predict( ... ). The cluster label will be zero if the model has been trained but no prediction has been run.

Returns
returns the most recent predicted cluster label

Definition at line 385 of file ClusterTree.cpp.

const ClusterTreeNode * GRT::ClusterTree::getTree ( ) const

Gets a pointer to the regression tree. NULL will be returned if the decision tree model has not be trained.

Returns
returns a const pointer to the regression tree

Definition at line 381 of file ClusterTree.cpp.

bool GRT::ClusterTree::loadModelFromFile ( fstream &  file)
virtual

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

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

Reimplemented from GRT::MLBase.

Definition at line 263 of file ClusterTree.cpp.

ClusterTree & GRT::ClusterTree::operator= ( const ClusterTree rhs)

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

Parameters
constClusterTreev &rhs: another instance of a ClusterTree
Returns
returns a pointer to this instance of the ClusterTree

Definition at line 65 of file ClusterTree.cpp.

bool GRT::ClusterTree::predict_ ( VectorDouble &  inputVector)
virtual

This predicts the class of the inputVector. This overrides the predict function in the Regressifier base class.

Parameters
VectorDoubleinputVector: the input vector to predict
Returns
returns true if the prediction was performed, false otherwise

Reimplemented from GRT::MLBase.

Definition at line 174 of file ClusterTree.cpp.

bool GRT::ClusterTree::print ( ) const
virtual

Prints the tree to std::cout.

Returns
returns true if the model was printed

Reimplemented from GRT::Tree.

Definition at line 221 of file ClusterTree.cpp.

bool GRT::ClusterTree::saveModelToFile ( fstream &  file) const
virtual

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

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

Reimplemented from GRT::MLBase.

Definition at line 227 of file ClusterTree.cpp.

bool GRT::ClusterTree::setMinRMSErrorPerNode ( const double  minRMSErrorPerNode)

Sets the minimum RMS error that needs to be exceeded for the tree to continue growing at a specific node.

Returns
returns true if the parameter was updated

Definition at line 393 of file ClusterTree.cpp.

bool GRT::ClusterTree::train_ ( MatrixDouble trainingData)
virtual

This trains the RegressionTree model, using the labelled regression data. This overrides the train function in the Regressifier base class.

Parameters
RegressionDatatrainingData: a reference to the training data
Returns
returns true if the RegressionTree model was trained, false otherwise

Reimplemented from GRT::Clusterer.

Definition at line 119 of file ClusterTree.cpp.


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