Train a linear model to whiten the data. More...

#include <shark/Algorithms/Trainers/NormalizeComponentsZCA.h>

Inheritance diagram for shark::NormalizeComponentsZCA:

Public Member Functions
	NormalizeComponentsZCA (double targetVariance=1.0)

std::string	name () const
	From INameable: return the class name.

void	train (ModelType &model, UnlabeledData< RealVector > const &input)

Public Member Functions inherited from shark::AbstractUnsupervisedTrainer< LinearModel< RealVector > >
virtual void	train (ModelType &model, DatasetType const &inputset)=0
	Core of the Trainer interface.

Public Member Functions inherited from shark::INameable
virtual	~INameable ()

Public Member Functions inherited from shark::ISerializable
virtual	~ISerializable ()
	Virtual d'tor.

virtual void	read (InArchive &archive)
	Read the component from the supplied archive.

virtual void	write (OutArchive &archive) const
	Write the component to the supplied archive.

void	load (InArchive &archive, unsigned int version)
	Versioned loading of components, calls read(...).

void	save (OutArchive &archive, unsigned int version) const
	Versioned storing of components, calls write(...).

	BOOST_SERIALIZATION_SPLIT_MEMBER ()

Public Attributes
double	m_targetVariance

Additional Inherited Members
Public Types inherited from shark::AbstractUnsupervisedTrainer< LinearModel< RealVector > >
typedef LinearModel< RealVector >	ModelType

typedef Model::InputType	InputType

typedef UnlabeledData< InputType >	DatasetType

Detailed Description

Train a linear model to whiten the data.

ZCA does whitening in the sense that it sets the mean to zero and the covariance to the Identity. However in contrast to NormalizeComponentsWhitening it makes sure that the initial and end coordinate system are the same and just rescales the data. The effect is, that image data still resembles images after applying ZCA in contrast to other methods which rotate the data randomly.

Definition at line 55 of file NormalizeComponentsZCA.h.

Constructor & Destructor Documentation

◆ NormalizeComponentsZCA()

shark::NormalizeComponentsZCA::NormalizeComponentsZCA ( double targetVariance = 1.0 )

inline

Definition at line 61 of file NormalizeComponentsZCA.h.

References m_targetVariance, and SHARK_RUNTIME_CHECK.

Member Function Documentation

◆ name()

std::string shark::NormalizeComponentsZCA::name ( ) const

inlinevirtual

From INameable: return the class name.

Reimplemented from shark::INameable.

Definition at line 67 of file NormalizeComponentsZCA.h.

◆ train()

void shark::NormalizeComponentsZCA::train	(	ModelType &	model,
		UnlabeledData< RealVector > const &	input
	)

inline

Definition at line 70 of file NormalizeComponentsZCA.h.

References shark::covariance(), shark::dataDimension(), m_targetVariance, shark::mean(), shark::meanvar(), shark::Data< Type >::numberOfElements(), shark::LinearModel< InputType, ActivationFunction >::setStructure(), and SHARK_RUNTIME_CHECK.

Member Data Documentation

◆ m_targetVariance

double shark::NormalizeComponentsZCA::m_targetVariance

Definition at line 60 of file NormalizeComponentsZCA.h.

Referenced by NormalizeComponentsZCA(), and train().

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

include/shark/Algorithms/Trainers/NormalizeComponentsZCA.h

Public Member Functions

Public Attributes

Additional Inherited Members