The Logistic class represents a logistic loss function. More...

#include <logistic.h>

Inheritance diagram for slope::Logistic:

Collaboration diagram for slope::Logistic:

Public Member Functions
double	loss (const Eigen::MatrixXd &eta, const Eigen::MatrixXd &y)
	Calculates the loss for the logistic loss function.

double	dual (const Eigen::MatrixXd &theta, const Eigen::MatrixXd &y, const Eigen::VectorXd &w)
	Calculates the dual for the logistic loss function.

Eigen::MatrixXd	residual (const Eigen::MatrixXd &eta, const Eigen::MatrixXd &y)
	Calculates the residual for the logistic loss function.

Eigen::MatrixXd	preprocessResponse (const Eigen::MatrixXd &y)
	Preprocesses the response for the quadratic model.

void	updateWeightsAndWorkingResponse (Eigen::VectorXd &w, Eigen::VectorXd &z, const Eigen::VectorXd &eta, const Eigen::VectorXd &y)
	Updates the weights and working response for the logistic loss function.

Eigen::MatrixXd	link (const Eigen::MatrixXd &mu)
	The link function.

Eigen::MatrixXd	inverseLink (const Eigen::MatrixXd &eta)
	The inverse link function, also known as the mean function.

Eigen::MatrixXd	predict (const Eigen::MatrixXd &eta)
	Return predicted response, that is 0 or 1 depending on the predicted probabilities.

Public Member Functions inherited from slope::Loss
virtual	~Loss ()=default
	Destructor for the Loss class.

virtual void	updateIntercept (Eigen::VectorXd &beta0, const Eigen::MatrixXd &eta, const Eigen::MatrixXd &y)
	Updates the intercept with a gradient descent update. Also updates the linear predictor (but not the residual).

virtual double	deviance (const Eigen::MatrixXd &eta, const Eigen::MatrixXd &y)
	Computes deviance, which is 2 times the difference between the loglikelihood of the model and the loglikelihood of the null (intercept-only) model.

Additional Inherited Members
Protected Member Functions inherited from slope::Loss
	Loss (double lipschitz_constant)
	Constructs an loss function with specified Lipschitz constant.

Detailed Description

The Logistic class represents a logistic loss function.

The logistic loss function is used for binary classification problems. It calculates the loss, dual, residual, and updates weights and working response.

Definition at line 18 of file logistic.h.

Constructor & Destructor Documentation

◆ Logistic()

slope::Logistic::Logistic ( )

inlineexplicit

Definition at line 21 of file logistic.h.

Member Function Documentation

◆ dual()

double slope::Logistic::dual	(	const Eigen::MatrixXd &	theta,
		const Eigen::MatrixXd &	y,
		const Eigen::VectorXd &	w
	)

virtual

Calculates the dual for the logistic loss function.

Parameters

theta	The dual variables.
y	The true labels.
w	Weights

Returns: The dual value.

Implements slope::Loss.

Definition at line 16 of file logistic.cpp.

◆ inverseLink()

Eigen::MatrixXd slope::Logistic::inverseLink ( const Eigen::MatrixXd & eta )

virtual

The inverse link function, also known as the mean function.

Parameters

eta Mean

Returns: \( \frac{1}{1 + \exp(-\eta)} \)

Implements slope::Loss.

Definition at line 71 of file logistic.cpp.

◆ link()

Eigen::MatrixXd slope::Logistic::link ( const Eigen::MatrixXd & mu )

virtual

The link function.

Parameters

mu Mean

Returns: \( \log(\frac{\mu}{1 - \mu}) \)

Implements slope::Loss.

Definition at line 63 of file logistic.cpp.

◆ loss()

double slope::Logistic::loss	(	const Eigen::MatrixXd &	eta,
		const Eigen::MatrixXd &	y
	)

virtual

Calculates the loss for the logistic loss function.

Parameters

eta	The predicted values.
y	The true labels.

Returns: The loss value.

Implements slope::Loss.

Definition at line 8 of file logistic.cpp.

◆ predict()

Eigen::MatrixXd slope::Logistic::predict ( const Eigen::MatrixXd & eta )

virtual

Return predicted response, that is 0 or 1 depending on the predicted probabilities.

Parameters

eta	The linear predictor

Returns: The predicted response

Implements slope::Loss.

Definition at line 78 of file logistic.cpp.

◆ preprocessResponse()

Eigen::MatrixXd slope::Logistic::preprocessResponse ( const Eigen::MatrixXd & y )

virtual

Preprocesses the response for the quadratic model.

Checks if the response is in {0, 1} and converts it otherwise

Parameters

y	Response vector (in {0,1})

Returns: Modified response.

Implements slope::Loss.

Definition at line 36 of file logistic.cpp.

◆ residual()

Eigen::MatrixXd slope::Logistic::residual	(	const Eigen::MatrixXd &	eta,
		const Eigen::MatrixXd &	y
	)

virtual

Calculates the residual for the logistic loss function.

Parameters

eta	The predicted values.
y	The true labels.

Returns: The residual vector.

Implements slope::Loss.

Definition at line 30 of file logistic.cpp.

◆ updateWeightsAndWorkingResponse()

void slope::Logistic::updateWeightsAndWorkingResponse	(	Eigen::VectorXd &	w,
		Eigen::VectorXd &	z,
		const Eigen::VectorXd &	eta,
		const Eigen::VectorXd &	y
	)

virtual

Updates the weights and working response for the logistic loss function.

Parameters

w	The weights.
z	The working response.
eta	The predicted values.
y	The true labels.

Implements slope::Loss.

Definition at line 50 of file logistic.cpp.

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

src/slope/losses/logistic.h
src/slope/losses/logistic.cpp

Public Member Functions

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

◆ Logistic()

Member Function Documentation

◆ dual()

◆ inverseLink()

◆ link()

◆ loss()

◆ predict()

◆ preprocessResponse()

◆ residual()

◆ updateWeightsAndWorkingResponse()