#include <RadialBasisFunction.h>

Inheritance diagram for RadialBasisFunction:

Public Member Functions
	RadialBasisFunction (UINT numSteps=100, Float positiveClassificationThreshold=0.9, Float minAlphaSearchRange=0.001, Float maxAlphaSearchRange=1.0)

virtual	~RadialBasisFunction ()

	RadialBasisFunction (const RadialBasisFunction &rhs)

RadialBasisFunction &	operator= (const RadialBasisFunction &rhs)

virtual bool	deepCopyFrom (const WeakClassifier *weakClassifer)

virtual bool	train (ClassificationData &trainingData, VectorFloat &weights)

virtual Float	predict (const VectorFloat &x)

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

virtual bool	loadModelFromFile (std::fstream &file)

virtual void	print () const

UINT	getNumSteps () const

Float	getPositiveClassificationThreshold () const

Float	getAlpha () const

Float	getMinAlphaSearchRange () const

Float	getMaxAlphaSearchRange () const

VectorFloat	getRBFCentre () const

Public Member Functions inherited from WeakClassifier
	WeakClassifier ()

virtual	~WeakClassifier ()

	WeakClassifier (const WeakClassifier &rhs)

WeakClassifier &	operator= (const WeakClassifier &rhs)

bool	copyBaseVariables (const WeakClassifier *weakClassifer)

virtual Float	getPositiveClassLabel () const

virtual Float	getNegativeClassLabel () const

std::string	getWeakClassifierType () const

bool	getTrained () const

UINT	getNumInputDimensions () const

bool	getTrainingLoggingEnabled () const

bool	setTrainingLoggingEnabled (const bool enabled)

WeakClassifier *	createNewInstance () const

Protected Member Functions
Float	rbf (const VectorFloat &a, const VectorFloat &b)

Protected Attributes
UINT	numSteps

Float	positiveClassificationThreshold

Float	alpha

Float	gamma

Float	minAlphaSearchRange

Float	maxAlphaSearchRange

VectorFloat	rbfCentre

Protected Attributes inherited from WeakClassifier
std::string	weakClassifierType
	A string that represents the weak classifier type, e.g. DecisionStump.

bool	trained
	A flag to show if the weak classifier model has been trained.

UINT	numInputDimensions
	The number of input dimensions to the weak classifier.

TrainingLog	trainingLog

ErrorLog	errorLog

WarningLog	warningLog

Static Protected Attributes
static RegisterWeakClassifierModule< RadialBasisFunction >	registerModule
	This is used to register the DecisionStump with the WeakClassifier base class.

Additional Inherited Members
Public Types inherited from WeakClassifier
typedef std::map< std::string, WeakClassifier ()() >	StringWeakClassifierMap

Static Public Member Functions inherited from WeakClassifier
static WeakClassifier *	createInstanceFromString (std::string const &weakClassifierType)

Static Protected Member Functions inherited from WeakClassifier
static StringWeakClassifierMap *	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.

Definition at line 40 of file RadialBasisFunction.h.

Constructor & Destructor Documentation

RadialBasisFunction::RadialBasisFunction	(	UINT	numSteps = `100`,
		Float	positiveClassificationThreshold = `0.9`,
		Float	minAlphaSearchRange = `0.001`,
		Float	maxAlphaSearchRange = `1.0`
	)

Default Constructor.

Sets the number of steps that will be used to search for the best alpha value during the training phase, in addition to other parameters used to control the RBF learning algorithm.

Parameters

numSteps	sets the number of steps that will be used to search for the best alpha value during the training phase. Default value = 100
positiveClassificationThreshold	sets the positive classification threshold, this parameter is the threshold that defines if a value is classified as a positive sample or a negative sample. Default value = 0.9
minAlphaSearchRange	the minimum value used to search for the best alpha value. Default value = 0.001
maxAlphaSearchRange	the maximum value used to search for the best alpha value. Default value = 1.0

Definition at line 36 of file RadialBasisFunction.cpp.

RadialBasisFunction::~RadialBasisFunction ( )

virtual

Default Destructor.

Definition at line 51 of file RadialBasisFunction.cpp.

RadialBasisFunction::RadialBasisFunction ( const RadialBasisFunction & rhs )

Default Copy Constructor. Defines how the data from the rhs GRT::RadialBasisFunction instance is copied to this GRT::RadialBasisFunction instance.

Definition at line 55 of file RadialBasisFunction.cpp.

Member Function Documentation

bool RadialBasisFunction::deepCopyFrom ( const WeakClassifier * weakClassifer )

virtual

This function enables the data from one GRT::RadialBasisFunction instance to be copied into this GRT::RadialBasisFunction instance.

Parameters

weakClassifer a pointer to the Classifier Base Class, this should be pointing to another GRT::RadialBasisFunction instance

Returns: returns true if the clone was successfull, false otherwise

Reimplemented from WeakClassifier.

Definition at line 73 of file RadialBasisFunction.cpp.

Float RadialBasisFunction::getAlpha ( ) const

Gets the current alpha value, this is used in the RBF. You can compute the RBF gamma parameter by: -1.0/(2.0*SQR(alpha)).

Returns: returns the alpha value

Definition at line 336 of file RadialBasisFunction.cpp.

Float RadialBasisFunction::getMaxAlphaSearchRange ( ) const

Gets the maxAlphaSearchRange value, this is the maximum value used to search for the best alpha value.

Returns: returns the maxAlphaSearchRange value

Definition at line 344 of file RadialBasisFunction.cpp.

Float RadialBasisFunction::getMinAlphaSearchRange ( ) const

Gets the minAlphaSearchRange value, this is the minimum value used to search for the best alpha value.

Returns: returns the minAlphaSearchRange value

Definition at line 340 of file RadialBasisFunction.cpp.

UINT RadialBasisFunction::getNumSteps ( ) const

This function gets the number of steps parameter which sets how many steps are used to search for the best RBF alpha values.

Returns: returns an UINT representing the numSteps parameter

Definition at line 328 of file RadialBasisFunction.cpp.

Float RadialBasisFunction::getPositiveClassificationThreshold ( ) const

This function gets the positiveClassificationThreshold, if the output of the RBF function is greater than or equal to the positiveClassificationThreshold then the input sample will be classified as a positive sample, otherwise it will be classified as a negative sample.

Returns: returns the positiveClassificationThreshold

Definition at line 332 of file RadialBasisFunction.cpp.

VectorFloat RadialBasisFunction::getRBFCentre ( ) const

Gets the RBF center.

Returns: returns the RBF centre.

Definition at line 324 of file RadialBasisFunction.cpp.

bool RadialBasisFunction::loadModelFromFile ( std::fstream & file )

virtual

This function loads an RBF model from a file.

fstream file: a reference to the file you want to load the RBF model from

Returns: returns true if the RBF was loaded successfull, false otherwise

Reimplemented from WeakClassifier.

Definition at line 231 of file RadialBasisFunction.cpp.

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

Defines how the data from the rhs GRT::RadialBasisFunction instance is copied to this GRT::RadialBasisFunction instance.

Definition at line 59 of file RadialBasisFunction.cpp.

Float RadialBasisFunction::predict ( const VectorFloat & x )

virtual

This function predicts the class label of the input vector, given the current RBF model. The class label returned will either be positive (WEAK_CLASSIFIER_POSITIVE_CLASS_LABEL) or negative (WEAK_CLASSIFIER_NEGATIVE_CLASS_LABEL).

Parameters

x	the vector used for prediction

Returns: returns the predicted class label, which will be either positive or negative

Reimplemented from WeakClassifier.

Definition at line 185 of file RadialBasisFunction.cpp.

void RadialBasisFunction::print ( ) const

virtual

This function prints out some basic info about the RBF to std::cout.

Reimplemented from WeakClassifier.

Definition at line 321 of file RadialBasisFunction.cpp.

bool RadialBasisFunction::saveModelToFile ( std::fstream & file ) const

virtual

This function saves the current RBF model to a file.

fstream file: a reference to the file you want to save the RBF model to

Returns: returns true if the RBF was saved successfull, false otherwise

Reimplemented from WeakClassifier.

Definition at line 200 of file RadialBasisFunction.cpp.

bool RadialBasisFunction::train	(	ClassificationData &	trainingData,
		VectorFloat &	weights
	)

virtual

This function trains the RBF, using the weighted labelled training data.

Parameters

trainingData	the labelled training data
weights	the corresponding weights for each sample in the labelled training data

Returns: returns true if the RBF was trained successfull, false otherwise

Reimplemented from WeakClassifier.

Definition at line 84 of file RadialBasisFunction.cpp.

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

Public Member Functions

Protected Member Functions

Protected Attributes

Static Protected Attributes

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation