Cost function interface. More...

#include <shark/ObjectiveFunctions/AbstractCost.h>

Inheritance diagram for shark::AbstractCost< LabelT, OutputT >:

Public Types
enum	Feature { HAS_FIRST_DERIVATIVE = 1 , HAS_SECOND_DERIVATIVE = 2 , IS_LOSS_FUNCTION = 4 }
	list of features a cost function can have More...

typedef OutputT	OutputType

typedef LabelT	LabelType

typedef Batch< OutputType >::type	BatchOutputType

typedef Batch< LabelType >::type	BatchLabelType

typedef TypedFlags< Feature >	Features

typedef TypedFeatureNotAvailableException< Feature >	FeatureNotAvailableException

Public Member Functions
virtual	~AbstractCost ()

const Features &	features () const

virtual void	updateFeatures ()

bool	hasFirstDerivative () const
	returns true when the first parameter derivative is implemented

bool	isLossFunction () const
	returns true when the cost function is in fact a loss function

virtual double	eval (Data< LabelType > const &targets, Data< OutputType > const &predictions) const =0

double	operator() (Data< LabelType > const &targets, Data< OutputType > const &predictions) const

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

virtual std::string	name () const
	returns the name of the object

Protected Attributes
Features	m_features

Detailed Description

template<class LabelT, class OutputT = LabelT>
class shark::AbstractCost< LabelT, OutputT >

Cost function interface.

: In Shark a cost function encodes the severity of a deviation of predictions from targets. This concept is more general than that or a loss function, because it does not necessarily amount to (uniformly) averaging a loss function over samples. In general, the loss depends on the true (training) label and the prediction in a not necessarily symmetric way. Also, in the most general case predictions can be in a different format than labels. E.g., the model prediction could be a probability distribution, while the label is a single value.

: The concept of an AbstractCost function is different from that encoded by the ErrorFunction class. A cost function compares model predictions to labels. It does not know about the model making the predictions, and thus it can not handle LabeledData directly. However, it is one of the components necessary to process LabeledData in an ErrorFunction.

Definition at line 73 of file AbstractCost.h.

Member Typedef Documentation

◆ BatchLabelType

template<class LabelT , class OutputT = LabelT>

typedef Batch<LabelType>::type shark::AbstractCost< LabelT, OutputT >::BatchLabelType

Definition at line 79 of file AbstractCost.h.

◆ BatchOutputType

template<class LabelT , class OutputT = LabelT>

typedef Batch<OutputType>::type shark::AbstractCost< LabelT, OutputT >::BatchOutputType

Definition at line 78 of file AbstractCost.h.

◆ FeatureNotAvailableException

template<class LabelT , class OutputT = LabelT>

typedef TypedFeatureNotAvailableException<Feature> shark::AbstractCost< LabelT, OutputT >::FeatureNotAvailableException

Definition at line 91 of file AbstractCost.h.

◆ Features

template<class LabelT , class OutputT = LabelT>

typedef TypedFlags<Feature> shark::AbstractCost< LabelT, OutputT >::Features

Definition at line 91 of file AbstractCost.h.

◆ LabelType

template<class LabelT , class OutputT = LabelT>

typedef LabelT shark::AbstractCost< LabelT, OutputT >::LabelType

Definition at line 77 of file AbstractCost.h.

◆ OutputType

template<class LabelT , class OutputT = LabelT>

typedef OutputT shark::AbstractCost< LabelT, OutputT >::OutputType

Definition at line 76 of file AbstractCost.h.

Member Enumeration Documentation

◆ Feature

template<class LabelT , class OutputT = LabelT>

enum shark::AbstractCost::Feature

list of features a cost function can have

Enumerator
HAS_FIRST_DERIVATIVE
HAS_SECOND_DERIVATIVE
IS_LOSS_FUNCTION

Definition at line 85 of file AbstractCost.h.

Constructor & Destructor Documentation

◆ ~AbstractCost()

template<class LabelT , class OutputT = LabelT>

virtual shark::AbstractCost< LabelT, OutputT >::~AbstractCost ( )

inlinevirtual

Definition at line 81 of file AbstractCost.h.

Member Function Documentation

◆ eval()

template<class LabelT , class OutputT = LabelT>

virtual double shark::AbstractCost< LabelT, OutputT >::eval	(	Data< LabelType > const &	targets,
		Data< OutputType > const &	predictions
	)		const

pure virtual

Evaluates the cost of predictions, given targets.

Parameters

targets	target values
predictions	predictions, typically made by a model

Implemented in shark::NegativeAUC< LabelType, OutputType >, shark::NegativeWilcoxonMannWhitneyStatistic< LabelType, OutputType >, shark::AbsoluteLoss< VectorType >, shark::AbstractLoss< LabelT, OutputT >, shark::SquaredLoss< OutputType, LabelType >, and shark::ZeroOneLoss< LabelType, OutputType >.

Referenced by shark::CrossValidationError< ModelTypeT, LabelTypeT >::eval(), and shark::AbstractCost< LabelT, OutputT >::operator()().

◆ features()

template<class LabelT , class OutputT = LabelT>

const Features & shark::AbstractCost< LabelT, OutputT >::features ( ) const

inline

Definition at line 91 of file AbstractCost.h.

◆ hasFirstDerivative()

template<class LabelT , class OutputT = LabelT>

bool shark::AbstractCost< LabelT, OutputT >::hasFirstDerivative ( ) const

inline

returns true when the first parameter derivative is implemented

Definition at line 94 of file AbstractCost.h.

References shark::AbstractCost< LabelT, OutputT >::HAS_FIRST_DERIVATIVE, and shark::AbstractCost< LabelT, OutputT >::m_features.

◆ isLossFunction()

template<class LabelT , class OutputT = LabelT>

bool shark::AbstractCost< LabelT, OutputT >::isLossFunction ( ) const

inline

returns true when the cost function is in fact a loss function

Definition at line 103 of file AbstractCost.h.

References shark::AbstractCost< LabelT, OutputT >::IS_LOSS_FUNCTION, and shark::AbstractCost< LabelT, OutputT >::m_features.

◆ operator()()

template<class LabelT , class OutputT = LabelT>

double shark::AbstractCost< LabelT, OutputT >::operator()	(	Data< LabelType > const &	targets,
		Data< OutputType > const &	predictions
	)		const

inline

Evaluates the cost of predictions, given targets.

Parameters

targets	Targets of the predictions
predictions	Predictions to be compared with the targets

Returns: The costs (e.g., error, regret)

Definition at line 116 of file AbstractCost.h.

References shark::AbstractCost< LabelT, OutputT >::eval().

◆ updateFeatures()

template<class LabelT , class OutputT = LabelT>

virtual void shark::AbstractCost< LabelT, OutputT >::updateFeatures ( )

inlinevirtual

Definition at line 91 of file AbstractCost.h.

Member Data Documentation

◆ m_features

template<class LabelT , class OutputT = LabelT>

Features shark::AbstractCost< LabelT, OutputT >::m_features

protected

Definition at line 91 of file AbstractCost.h.

Referenced by shark::AbstractLoss< LabelT, OutputT >::AbstractLoss(), shark::EpsilonHingeLoss::EpsilonHingeLoss(), shark::AbstractCost< LabelT, OutputT >::hasFirstDerivative(), shark::HingeLoss::HingeLoss(), shark::HuberLoss::HuberLoss(), shark::AbstractCost< LabelT, OutputT >::isLossFunction(), shark::SquaredEpsilonHingeLoss::SquaredEpsilonHingeLoss(), shark::SquaredHingeLoss::SquaredHingeLoss(), shark::SquaredLoss< OutputType, LabelType >::SquaredLoss(), shark::SquaredLoss< OutputType, unsigned int >::SquaredLoss(), and shark::SquaredLoss< Sequence, Sequence >::SquaredLoss().

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

include/shark/ObjectiveFunctions/AbstractCost.h

Public Types

Public Member Functions

Protected Attributes

Detailed Description

Member Typedef Documentation

◆ BatchLabelType

◆ BatchOutputType

◆ FeatureNotAvailableException

◆ Features

◆ LabelType

◆ OutputType

Member Enumeration Documentation

◆ Feature

Constructor & Destructor Documentation

◆ ~AbstractCost()

Member Function Documentation

◆ eval()

◆ features()

◆ hasFirstDerivative()

◆ isLossFunction()

◆ operator()()

◆ updateFeatures()

Member Data Documentation

◆ m_features