File probability_calibration_isotonic.hpp

namespace boosting

class IIsotonicMarginalProbabilityCalibratorConfig

#include <probability_calibration_isotonic.hpp>

Defines an interface for all classes that allow to configure a calibrator that fits a model for the calibration of marginal probabilities via isotonic regression.

Subclassed by boosting::IsotonicMarginalProbabilityCalibratorConfig

Public Functions

inline virtual ~IIsotonicMarginalProbabilityCalibratorConfig()

virtual bool isHoldoutSetUsed() const = 0

Returns whether the calibration model is fit to the examples in the holdout set, if available, or not.

Returns:

True, if the calibration model is fit to the examples in the holdout set, if available, false if the training set is used instead

virtual IIsotonicMarginalProbabilityCalibratorConfig &setUseHoldoutSet(bool useHoldoutSet) = 0

Sets whether the calibration model should be fit to the examples in the holdout set, if available, or not.

Parameters:

useHoldoutSet – True, if the calibration model should be fit to the examples in the holdout set, if available, false if the training set should be used instead

Returns:

A reference to an object of type IIsotonicMarginalProbabilityCalibratorConfig that allows further configuration of the calibrator

class IsotonicMarginalProbabilityCalibratorConfig : public boosting::IIsotonicMarginalProbabilityCalibratorConfig, public IMarginalProbabilityCalibratorConfig

#include <probability_calibration_isotonic.hpp>

Allows to configure a calibrator that fits a model for the calibration of marginal probabilities via isotonic regression.

Public Functions

IsotonicMarginalProbabilityCalibratorConfig(const std::unique_ptr<ILossConfig> &lossConfigPtr)

Parameters:

lossConfigPtr – A reference to an unique pointer that stores the configuration of the loss function

virtual bool isHoldoutSetUsed() const override

Returns whether the calibration model is fit to the examples in the holdout set, if available, or not.

Returns:

True, if the calibration model is fit to the examples in the holdout set, if available, false if the training set is used instead

virtual IIsotonicMarginalProbabilityCalibratorConfig &setUseHoldoutSet(bool useHoldoutSet) override

Sets whether the calibration model should be fit to the examples in the holdout set, if available, or not.

Parameters:

useHoldoutSet – True, if the calibration model should be fit to the examples in the holdout set, if available, false if the training set should be used instead

Returns:

A reference to an object of type IIsotonicMarginalProbabilityCalibratorConfig that allows further configuration of the calibrator

bool shouldUseHoldoutSet() const override

See also

IMarginalProbabilityCalibratorConfig::shouldUseHoldoutSet

std::unique_ptr<IMarginalProbabilityCalibratorFactory> createMarginalProbabilityCalibratorFactory() const override

See also

IMarginalProbabilityCalibratorConfig::createMarginalProbabilityCalibratorFactory

Private Members

bool useHoldoutSet_

const std::unique_ptr<ILossConfig> &lossConfigPtr_

class IIsotonicJointProbabilityCalibratorConfig

#include <probability_calibration_isotonic.hpp>

Defines an interface for all classes that allow to configure a calibrator that fits a model for the calibration of joint probabilities via isotonic regression.

Subclassed by boosting::IsotonicJointProbabilityCalibratorConfig

Public Functions

inline virtual ~IIsotonicJointProbabilityCalibratorConfig()

virtual bool isHoldoutSetUsed() const = 0

Returns whether the calibration model is fit to the examples in the holdout set, if available, or not.

Returns:

True, if the calibration model is fit to the examples in the holdout set, if available, false if the training set is used instead

virtual IIsotonicJointProbabilityCalibratorConfig &setUseHoldoutSet(bool useHoldoutSet) = 0

Sets whether the calibration model should be fit to the examples in the holdout set, if available, or not.

Parameters:

useHoldoutSet – True, if the calibration model should be fit to the examples in the holdout set, if available, false if the training set should be used instead

Returns:

A reference to an object of type IIsotonicJointProbabilityCalibratorConfig that allows further configuration of the calibrator

class IsotonicJointProbabilityCalibratorConfig : public boosting::IIsotonicJointProbabilityCalibratorConfig, public IJointProbabilityCalibratorConfig

#include <probability_calibration_isotonic.hpp>

Allows to configure a calibrator that fits a model for the calibration of joint probabilities via isotonic regression.

Public Functions

IsotonicJointProbabilityCalibratorConfig(const std::unique_ptr<ILossConfig> &lossConfigPtr)

Parameters:

lossConfigPtr – A reference to an unique pointer that stores the configuration of the loss function

virtual bool isHoldoutSetUsed() const override

Returns whether the calibration model is fit to the examples in the holdout set, if available, or not.

Returns:

True, if the calibration model is fit to the examples in the holdout set, if available, false if the training set is used instead

virtual IIsotonicJointProbabilityCalibratorConfig &setUseHoldoutSet(bool useHoldoutSet) override

Sets whether the calibration model should be fit to the examples in the holdout set, if available, or not.

Parameters:

useHoldoutSet – True, if the calibration model should be fit to the examples in the holdout set, if available, false if the training set should be used instead

Returns:

A reference to an object of type IIsotonicJointProbabilityCalibratorConfig that allows further configuration of the calibrator

bool shouldUseHoldoutSet() const override

See also

IJointProbabilityCalibratorConfig::shouldUseHoldoutSet

bool isLabelVectorSetNeeded() const override

See also

IJointProbabilityCalibratorConfig::isLabelVectorSetNeeeded

std::unique_ptr<IJointProbabilityCalibratorFactory> createJointProbabilityCalibratorFactory() const override

See also

IJointProbabilityCalibratorConfig::createJointProbabilityCalibratorFactory

Private Members

bool useHoldoutSet_

const std::unique_ptr<ILossConfig> &lossConfigPtr_