Support Vector Machines (Support Vector Machines (SVM)) More...

#include "generic.h"
#include "svmdataset.h"

Include dependency graph for svm.h:

This graph shows which files directly or indirectly include this file:

Classes
struct	VlSvmStatistics_
	SVM statistics This structure contains statistics characterising the state of the SVM solver, such as the current value of the objective function. More...
Typedefs
typedef struct VlSvm_	VlSvm
	SVM solver. This object implements VLFeat SVM solvers (see svm.h).
typedef struct VlSvmStatistics_	VlSvmStatistics
	SVM statistics This structure contains statistics characterising the state of the SVM solver, such as the current value of the objective function.
Enumerations
enum	VlSvmLossType { VlSvmLossHinge = 0, VlSvmLossHinge2 = 1, VlSvmLossL1, VlSvmLossL2, VlSvmLossLogistic }
	Type of SVM loss. More...
enum	VlSvmSolverStatus { VlSvmStatusTraining = 1, VlSvmStatusConverged, VlSvmStatusMaxNumIterationsReached }
	Solver status. More...
enum	VlSvmSolverType { VlSvmSolverNone = 0, VlSvmSolverSgd = 1, VlSvmSolverSdca }
	Type of SVM solver. More...
Functions
Create and destroy
VL_EXPORT VlSvm *	vl_svm_new (VlSvmSolverType type, double const data, vl_size dimension, vl_size numData, double const labels, double lambda)
	Create a new object with plain data.
VL_EXPORT VlSvm *	vl_svm_new_with_dataset (VlSvmSolverType type, VlSvmDataset dataset, double const labels, double lambda)
	Create a new object with a dataset.
VL_EXPORT VlSvm *	vl_svm_new_with_abstract_data (VlSvmSolverType type, void data, vl_size dimension, vl_size numData, double const labels, double lambda)
	Create a new object with abstract data.
VL_EXPORT void	vl_svm_delete (VlSvm *self)
	Delete object.
Retrieve parameters and data
VL_EXPORT VlSvmStatistics const *	vl_svm_get_statistics (VlSvm const *self)
	Get the solver statistics.
VL_EXPORT double const *	vl_svm_get_model (VlSvm const *self)
	Get the SVM model.
VL_EXPORT double	vl_svm_get_bias (VlSvm const *self)
	Get the value of the bias.
VL_EXPORT vl_size	vl_svm_get_dimension (VlSvm *self)
	Get the model dimenison.
VL_EXPORT vl_size	vl_svm_get_num_data (VlSvm *self)
	Get the number of data samples.
VL_EXPORT double	vl_svm_get_epsilon (VlSvm const *self)
	Get the convergence threshold.
VL_EXPORT double	vl_svm_get_bias_learning_rate (VlSvm const *self)
	Get the bias leraning rate.
VL_EXPORT vl_size	vl_svm_get_max_num_iterations (VlSvm const *self)
	Get the maximum number of iterations.
VL_EXPORT vl_size	vl_svm_get_diagnostic_frequency (VlSvm const *self)
	Get the diagnostic frequency.
VL_EXPORT VlSvmSolverType	vl_svm_get_solver (VlSvm const *self)
	Get the SVM solver type.
VL_EXPORT double	vl_svm_get_bias_multiplier (VlSvm const *self)
	Get the bias multiplier.
VL_EXPORT double	vl_svm_get_lambda (VlSvm const *self)
	Get the regularizer parameter lambda.
VL_EXPORT vl_size	vl_svm_get_iteration_number (VlSvm const *self)
	Get the current iteration number.
VL_EXPORT double const *	vl_svm_get_scores (VlSvm const *self)
	Get the scores of the data points.
VL_EXPORT double const *	vl_svm_get_weights (VlSvm const *self)
	Get the data weights.
Set parameters
VL_EXPORT void	vl_svm_set_epsilon (VlSvm *self, double epsilon)
	Set the convergence threshold.
VL_EXPORT void	vl_svm_set_bias_learning_rate (VlSvm *self, double rate)
	Set the bias learning rate.
VL_EXPORT void	vl_svm_set_max_num_iterations (VlSvm *self, vl_size maxNumIterations)
	Set the maximum number of iterations.
VL_EXPORT void	vl_svm_set_diagnostic_frequency (VlSvm *self, vl_size f)
	Set the diagnostic frequency.
VL_EXPORT void	vl_svm_set_bias_multiplier (VlSvm *self, double b)
	Set the bias multiplier.
VL_EXPORT void	vl_svm_set_model (VlSvm self, double const model)
	Set the SVM model.
VL_EXPORT void	vl_svm_set_bias (VlSvm *self, double b)
	Set the SVM bias.
VL_EXPORT void	vl_svm_set_iteration_number (VlSvm *self, vl_uindex n)
	Set the current iteratio number.
VL_EXPORT void	vl_svm_set_weights (VlSvm self, double const weights)
	Set the data weights.
VL_EXPORT void	vl_svm_set_diagnostic_function (VlSvm self, VlSvmDiagnosticFunction f, void data)
	Set the diagnostic function callback.
VL_EXPORT void	vl_svm_set_loss_function (VlSvm *self, VlSvmLossFunction f)
	Set the loss function callback.
VL_EXPORT void	vl_svm_set_loss_derivative_function (VlSvm *self, VlSvmLossFunction f)
	Set the loss derivative function callback.
VL_EXPORT void	vl_svm_set_conjugate_loss_function (VlSvm *self, VlSvmLossFunction f)
	Set the conjugate loss function callback.
VL_EXPORT void	vl_svm_set_dca_update_function (VlSvm *self, VlSvmDcaUpdateFunction f)
	Set the DCA update function callback.
VL_EXPORT void	vl_svm_set_data_functions (VlSvm *self, VlSvmInnerProductFunction inner, VlSvmAccumulateFunction acc)
	Set the data functions.
VL_EXPORT void	vl_svm_set_loss (VlSvm *self, VlSvmLossType loss)
	Set the loss function to one of the default types.
Process data
VL_EXPORT void	vl_svm_train (VlSvm *self)
	Run the SVM solver.
Loss functions
See also: Advanced SVM topics
VL_EXPORT double	vl_svm_hinge_loss (double label, double inner)
	SVM hinge loss.
VL_EXPORT double	vl_svm_hinge_loss_derivative (double label, double inner)
	SVM hinge loss derivative.
VL_EXPORT double	vl_svm_hinge_conjugate_loss (double label, double u)
	SVM hinge loss conjugate.
VL_EXPORT double	vl_svm_hinge_dca_update (double alpha, double inner, double norm2, double label)
	SVM hinge loss DCA update.
VL_EXPORT double	vl_svm_hinge2_loss (double label, double inner)
	SVM square hinge loss.
VL_EXPORT double	vl_svm_hinge2_loss_derivative (double label, double inner)
	SVM square hinge loss derivative.
VL_EXPORT double	vl_svm_hinge2_conjugate_loss (double label, double u)
	SVM square hinge loss conjugate.
VL_EXPORT double	vl_svm_hinge2_dca_update (double alpha, double inner, double norm2, double label)
	SVM square hinge loss DCA update.
VL_EXPORT double	vl_svm_l1_loss (double label, double inner)
	SVM l1 loss.
VL_EXPORT double	vl_svm_l1_loss_derivative (double label, double inner)
	SVM l1 loss derivative.
VL_EXPORT double	vl_svm_l1_conjugate_loss (double label, double u)
	SVM l1 loss conjugate.
VL_EXPORT double	vl_svm_l1_dca_update (double alpha, double inner, double norm2, double label)
	SVM l1 loss DCA update.
VL_EXPORT double	vl_svm_l2_loss (double label, double inner)
	SVM l2 loss.
VL_EXPORT double	vl_svm_l2_loss_derivative (double label, double inner)
	SVM l2 loss derivative.
VL_EXPORT double	vl_svm_l2_conjugate_loss (double label, double u)
	SVM l2 loss conjugate.
VL_EXPORT double	vl_svm_l2_dca_update (double alpha, double inner, double norm2, double label)
	SVM l2 loss DCA update.
VL_EXPORT double	vl_svm_logistic_loss (double label, double inner)
	SVM l2 loss.
VL_EXPORT double	vl_svm_logistic_loss_derivative (double label, double inner)
	SVM l2 loss derivative.
VL_EXPORT double	vl_svm_logistic_conjugate_loss (double label, double u)
	SVM l2 loss conjugate.
VL_EXPORT double	vl_svm_logistic_dca_update (double alpha, double inner, double norm2, double label)
	SVM l2 loss DCA update.

Detailed Description

Support Vector Machines (Support Vector Machines (SVM))

See also:: Support Vector Machines (SVM).

Author:: Milan Sulc; Daniele Perrone; Andrea Vedaldi

Definition in file svm.h.

Typedef Documentation

VlSvm

SVM solver. This object implements VLFeat SVM solvers (see svm.h).

Definition at line 32 of file svm.h.

typedef struct VlSvmStatistics_ VlSvmStatistics

SVM statistics This structure contains statistics characterising the state of the SVM solver, such as the current value of the objective function.

Not all fields are used by all solvers.

Enumeration Type Documentation

enum VlSvmLossType

Type of SVM loss.

Default SVM loss types. The loss can be set by using vl_svm_set_loss. Note that custom losses can be used too by using vl_svm_set_loss_function, vl_svm_set_loss_derivative_function, etc.

See also:: svm-loss-functions

Enumerator:

VlSvmLossHinge	Standard hinge loss.
VlSvmLossHinge2	Hinge loss squared.
VlSvmLossL1	L1 loss.
VlSvmLossL2	L2 loss.
VlSvmLossLogistic	Logistic loss.

Definition at line 53 of file svm.h.

enum VlSvmSolverStatus

Solver status.

Enumerator:

VlSvmStatusTraining	Optimization in progress.
VlSvmStatusConverged	Optimization finished because the convergence criterion was met.
VlSvmStatusMaxNumIterationsReached	Optimization finished without convergence.

Definition at line 63 of file svm.h.

enum VlSvmSolverType

Type of SVM solver.

Enumerator:

VlSvmSolverNone	No solver (used to evaluate an SVM).
VlSvmSolverSgd	SGD algorithm (Stochastic Gradient Descent).
VlSvmSolverSdca	SDCA algorithm (Stochastic Dual Coordinate Ascent).

Definition at line 38 of file svm.h.

Function Documentation

VL_EXPORT void vl_svm_delete ( VlSvm * self )

Delete object.

Parameters:

self object.

See also:: vl_svm_new

Definition at line 1097 of file svm.c.

VL_EXPORT double vl_svm_get_bias ( VlSvm const * self )

Get the value of the bias.

Parameters:

self object.

Returns:: bias $b$.

The value of the bias returned already include the effect of bias mutliplier.

Definition at line 1422 of file svm.c.

VL_EXPORT double vl_svm_get_bias_learning_rate ( VlSvm const * self )

Get the bias leraning rate.

Parameters:

self object

Returns:: bias learning rate.

Definition at line 1160 of file svm.c.

VL_EXPORT double vl_svm_get_bias_multiplier ( VlSvm const * self )

Get the bias multiplier.

Parameters:

self object.

Returns:: bias multiplier.

Definition at line 1186 of file svm.c.

VL_EXPORT vl_size vl_svm_get_diagnostic_frequency ( VlSvm const * self )

Get the diagnostic frequency.

Parameters:

self object.

Returns:: diagnostic frequency.

Definition at line 1260 of file svm.c.

VL_EXPORT vl_size vl_svm_get_dimension ( VlSvm * self )

Get the model dimenison.

Parameters:

self object.

Returns:: model dimension.

This is the dimensionality of the weight vector $$.

Definition at line 1348 of file svm.c.

VL_EXPORT double vl_svm_get_epsilon ( VlSvm const * self )

Get the convergence threshold.

Parameters:

self object

Returns:: epsilon threshold.

Definition at line 1135 of file svm.c.

VL_EXPORT vl_size vl_svm_get_iteration_number ( VlSvm const * self )

Get the current iteration number.

Parameters:

self object.

Returns:: current iteration number.

Definition at line 1212 of file svm.c.

VL_EXPORT double vl_svm_get_lambda ( VlSvm const * self )

Get the regularizer parameter lambda.

Parameters:

self object.

Returns:: diagnostic frequency.

Definition at line 1300 of file svm.c.

VL_EXPORT vl_size vl_svm_get_max_num_iterations ( VlSvm const * self )

Get the maximum number of iterations.

Parameters:

self object.

Returns:: maximum number of iterations.

Definition at line 1234 of file svm.c.

VL_EXPORT double const* vl_svm_get_model ( VlSvm const * self )

Get the SVM model.

Parameters:

self object.

Returns:: model.

This is the weight vector $$.

Definition at line 1374 of file svm.c.

VL_EXPORT vl_size vl_svm_get_num_data ( VlSvm * self )

Get the number of data samples.

Parameters:

self object.

Returns:: model number of data samples

This is the dimensionality of the weight vector $$.

Definition at line 1361 of file svm.c.

VL_EXPORT double const* vl_svm_get_scores ( VlSvm const * self )

Get the scores of the data points.

Parameters:

self object.

Returns:: vector of scores.

After training or during the diagnostic callback, this function can be used to retrieve the scores of the points, i.e. $ , + b$.

Definition at line 1448 of file svm.c.

VL_EXPORT VlSvmSolverType vl_svm_get_solver ( VlSvm const * self )

Get the SVM solver type.

Parameters:

self object.

Returns:: SVM solver type.

Definition at line 1271 of file svm.c.

VL_EXPORT VlSvmStatistics const* vl_svm_get_statistics ( VlSvm const * self )

Get the solver statistics.

Parameters:

self object.

Returns:: statistics.

Definition at line 1433 of file svm.c.

VL_EXPORT double const* vl_svm_get_weights ( VlSvm const * self )

Get the data weights.

Parameters:

self object.

Returns:: data weights.

Definition at line 1331 of file svm.c.

VL_EXPORT double vl_svm_hinge2_conjugate_loss	(	double	u,
		double	label
	)

SVM square hinge loss conjugate.

Parameters:

u	dual variable.
label	label value.

Returns:: conjugate loss.

Definition at line 1667 of file svm.c.

VL_EXPORT double vl_svm_hinge2_dca_update	(	double	alpha,
		double	inner,
		double	norm2,
		double	label
	)

SVM square hinge loss DCA update.

Parameters:

alpha	current value of the dual variable.
inner	inner product $^ $ of the sample with the SVM model.
norm2	normalization factor $\\|\\|^2/ n$.
label	label $y$ of the sample.

Returns:: incremental update $$ of the dual variable.

See also:: Stochastic Dual Coordinate Ascent

Definition at line 1678 of file svm.c.

VL_EXPORT double vl_svm_hinge2_loss	(	double	inner,
		double	label
	)

SVM square hinge loss.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1646 of file svm.c.

VL_EXPORT double vl_svm_hinge2_loss_derivative	(	double	inner,
		double	label
	)

SVM square hinge loss derivative.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1655 of file svm.c.

VL_EXPORT double vl_svm_hinge_conjugate_loss	(	double	u,
		double	label
	)

SVM hinge loss conjugate.

Parameters:

u	dual variable.
label	label value.

Returns:: conjugate loss.

Definition at line 1626 of file svm.c.

VL_EXPORT double vl_svm_hinge_dca_update	(	double	alpha,
		double	inner,
		double	norm2,
		double	label
	)

SVM hinge loss DCA update.

Parameters:

alpha	current value of the dual variable.
inner	inner product $^ $ of the sample with the SVM model.
norm2	normalization factor $\\|\\|^2/ n$.
label	label $y$ of the sample.

Returns:: incremental update $$ of the dual variable.

See also:: Stochastic Dual Coordinate Ascent

Definition at line 1638 of file svm.c.

VL_EXPORT double vl_svm_hinge_loss	(	double	inner,
		double	label
	)

SVM hinge loss.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1603 of file svm.c.

VL_EXPORT double vl_svm_hinge_loss_derivative	(	double	inner,
		double	label
	)

SVM hinge loss derivative.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1611 of file svm.c.

VL_EXPORT double vl_svm_l1_conjugate_loss	(	double	u,
		double	label
	)

SVM l1 loss conjugate.

Parameters:

u	dual variable.
label	label value.

Returns:: conjugate loss.

Definition at line 1706 of file svm.c.

VL_EXPORT double vl_svm_l1_dca_update	(	double	alpha,
		double	inner,
		double	norm2,
		double	label
	)

SVM l1 loss DCA update.

Parameters:

alpha	current value of the dual variable.
inner	inner product $^ $ of the sample with the SVM model.
norm2	normalization factor $\\|\\|^2/ n$.
label	label $y$ of the sample.

Returns:: incremental update $$ of the dual variable.

See also:: Stochastic Dual Coordinate Ascent

Definition at line 1717 of file svm.c.

VL_EXPORT double vl_svm_l1_loss	(	double	inner,
		double	label
	)

SVM l1 loss.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1686 of file svm.c.

VL_EXPORT double vl_svm_l1_loss_derivative	(	double	inner,
		double	label
	)

SVM l1 loss derivative.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1694 of file svm.c.

VL_EXPORT double vl_svm_l2_conjugate_loss	(	double	u,
		double	label
	)

SVM l2 loss conjugate.

Parameters:

u	dual variable.
label	label value.

Returns:: conjugate loss.

Definition at line 1746 of file svm.c.

VL_EXPORT double vl_svm_l2_dca_update	(	double	alpha,
		double	inner,
		double	norm2,
		double	label
	)

SVM l2 loss DCA update.

Parameters:

alpha	current value of the dual variable.
inner	inner product $^ $ of the sample with the SVM model.
norm2	normalization factor $\\|\\|^2/ n$.
label	label $y$ of the sample.

Returns:: incremental update $$ of the dual variable.

See also:: Stochastic Dual Coordinate Ascent

Definition at line 1753 of file svm.c.

VL_EXPORT double vl_svm_l2_loss	(	double	inner,
		double	label
	)

SVM l2 loss.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1729 of file svm.c.

VL_EXPORT double vl_svm_l2_loss_derivative	(	double	inner,
		double	label
	)

SVM l2 loss derivative.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1738 of file svm.c.

VL_EXPORT double vl_svm_logistic_conjugate_loss	(	double	u,
		double	label
	)

SVM l2 loss conjugate.

Parameters:

u	dual variable.
label	label value.

Returns:: conjugate loss.

Definition at line 1789 of file svm.c.

VL_EXPORT double vl_svm_logistic_dca_update	(	double	alpha,
		double	inner,
		double	norm2,
		double	label
	)

SVM l2 loss DCA update.

Parameters:

alpha	current value of the dual variable.
inner	inner product $^ $ of the sample with the SVM model.
norm2	normalization factor $\\|\\|^2/ n$.
label	label $y$ of the sample.

Returns:: incremental update $$ of the dual variable.

See also:: Stochastic Dual Coordinate Ascent

Definition at line 1801 of file svm.c.

VL_EXPORT double vl_svm_logistic_loss	(	double	inner,
		double	label
	)

SVM l2 loss.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1760 of file svm.c.

VL_EXPORT double vl_svm_logistic_loss_derivative	(	double	inner,
		double	label
	)

SVM l2 loss derivative.

Parameters:

inner	inner product between sample and model $^ $.
label	sample label $y$.

Returns:: value of the loss.

The interface is the same for a loss function, its derivative, or the conjugate loss.

See also:: SVM fundamentals

Definition at line 1773 of file svm.c.

VL_EXPORT VlSvm* vl_svm_new	(	VlSvmSolverType	type,
		double const *	data,
		vl_size	dimension,
		vl_size	numData,
		double const *	labels,
		double	lambda
	)

Create a new object with plain data.

Parameters:

type	type of SMV solver.
data	a pointer to a matrix of data.
dimension	dimension of the SVM model.
numData	number of training samples.
labels	training labels.
lambda	regularizer parameter.

Returns:: the new object.

data has one column per sample, in double format. More advanced inputs can be used with vl_svm_new_with_dataset and vl_svm_new_with_abstract_data.

See also:: vl_svm_delete

Definition at line 957 of file svm.c.

VL_EXPORT VlSvm* vl_svm_new_with_abstract_data	(	VlSvmSolverType	solver,
		void *	data,
		vl_size	dimension,
		vl_size	numData,
		double const *	labels,
		double	lambda
	)

Create a new object with abstract data.

Parameters:

solver	type of SMV solver.
data	pointer to the data.
dimension	dimension of the SVM model.
numData	num training samples.
labels	training samples labels.
lambda	regularizer parameter.

Returns:: the new object.

After calling this function, vl_svm_set_data_functions *must* be used to setup suitable callbacks for the inner product and accumulation operations (

See also:: svm-data-abstraction).; vl_svm_delete

Definition at line 1014 of file svm.c.

VL_EXPORT VlSvm* vl_svm_new_with_dataset	(	VlSvmSolverType	solver,
		VlSvmDataset *	dataset,
		double const *	labels,
		double	lambda
	)

Create a new object with a dataset.

Parameters:

solver	type of SMV solver.
dataset	SVM dataset object
labels	training samples labels.
lambda	regularizer parameter.

Returns:: the new object.

See also:: vl_svm_delete

Definition at line 980 of file svm.c.

VL_EXPORT void vl_svm_set_bias	(	VlSvm *	self,
		double	b
	)

Set the SVM bias.

Parameters:

self	object.
b	bias.

The function set the internal representation of the SVM bias to be equal to b (the bias multiplier is applied). The same remark that applies to vl_svm_set_model applies here too.

Definition at line 1406 of file svm.c.

VL_EXPORT void vl_svm_set_bias_learning_rate	(	VlSvm *	self,
		double	rate
	)

Set the bias learning rate.

Parameters:

self	object
rate	bias learning rate (positive).

This parameter applies only to the SGD solver.

Definition at line 1148 of file svm.c.

VL_EXPORT void vl_svm_set_bias_multiplier	(	VlSvm *	self,
		double	b
	)

Set the bias multiplier.

Parameters:

self	object
b	bias multiplier.

The *bias multiplier* is the value of the constant feature appended to the data vectors to implement the bias (Adding a bias).

Definition at line 1174 of file svm.c.

VL_EXPORT void vl_svm_set_conjugate_loss_function	(	VlSvm *	self,
		VlSvmLossFunction	f
	)

Set the conjugate loss function callback.

Parameters:

self	object.
f	loss function callback.

Note that setting up a loss requires specifying more than just one callback. See Loss functions for details.

.

Definition at line 1529 of file svm.c.

VL_EXPORT void vl_svm_set_data_functions	(	VlSvm *	self,
		VlSvmInnerProductFunction	inner,
		VlSvmAccumulateFunction	acc
	)

Set the data functions.

Parameters:

self	object.
inner	inner product function.
acc	accumulate function.

See svm-data-abstraction.

Definition at line 1492 of file svm.c.

VL_EXPORT void vl_svm_set_dca_update_function	(	VlSvm *	self,
		VlSvmDcaUpdateFunction	f
	)

Set the DCA update function callback.

Parameters:

self	object.
f	loss function callback.

Note that setting up a loss requires specifying more than just one callback. See Loss functions for details.

.

Definition at line 1539 of file svm.c.

VL_EXPORT void vl_svm_set_diagnostic_frequency	(	VlSvm *	self,
		vl_size	f
	)

Set the diagnostic frequency.

Parameters:

self	object.
f	diagnostic frequency (`>=` 1).

A diagnostic round (to test for convergence and to printout information) is performed every f iterations.

Definition at line 1248 of file svm.c.

VL_EXPORT void vl_svm_set_diagnostic_function	(	VlSvm *	self,
		VlSvmDiagnosticFunction	f,
		void *	data
	)

Set the diagnostic function callback.

Parameters:

self	object.
f	diagnostic function pointer.
data	pointer to data used by the diagnostic function.

Definition at line 1479 of file svm.c.

VL_EXPORT void vl_svm_set_epsilon	(	VlSvm *	self,
		double	epsilon
	)

Set the convergence threshold.

Parameters:

self	object
epsilon	threshold (non-negative).

Definition at line 1123 of file svm.c.

VL_EXPORT void vl_svm_set_iteration_number	(	VlSvm *	self,
		vl_uindex	n
	)

Set the current iteratio number.

Parameters:

self	object.
n	iteration number.

If called before training, this can be used with SGD for a warm start, as the net effect is to slow down the learning rate.

Definition at line 1201 of file svm.c.

VL_EXPORT void vl_svm_set_loss	(	VlSvm *	self,
		VlSvmLossType	loss
	)

Set the loss function to one of the default types.

Parameters:

self	object.
loss	type of loss function.

See also:: Loss functions.

Definition at line 1552 of file svm.c.

VL_EXPORT void vl_svm_set_loss_derivative_function	(	VlSvm *	self,
		VlSvmLossFunction	f
	)

Set the loss derivative function callback.

Parameters:

self	object.
f	loss function callback.

Note that setting up a loss requires specifying more than just one callback. See Loss functions for details.

.

Definition at line 1519 of file svm.c.

VL_EXPORT void vl_svm_set_loss_function	(	VlSvm *	self,
		VlSvmLossFunction	f
	)

Set the loss function callback.

Parameters:

self	object.
f	loss function callback.

Note that setting up a loss requires specifying more than just one callback. See Loss functions for details.

Definition at line 1509 of file svm.c.

VL_EXPORT void vl_svm_set_max_num_iterations	(	VlSvm *	self,
		vl_size	n
	)

Set the maximum number of iterations.

Parameters:

self	object.
n	maximum number of iterations.

Definition at line 1223 of file svm.c.

VL_EXPORT void vl_svm_set_model	(	VlSvm *	self,
		double const *	model
	)

Set the SVM model.

Parameters:

self	object.
model	model.

The function *copies* the content of the vector model to the internal model buffer. This operation can be used for warm start with the SGD algorithm, but has undefined effect with the SDCA algorithm.

Definition at line 1389 of file svm.c.

VL_EXPORT void vl_svm_set_weights	(	VlSvm *	self,
		double const *	weights
	)

Set the data weights.

Parameters:

self	object.
weights	data weights.

weights must be an array of non-negative weights. The loss of each data point is multiplied by the corresponding weight.

Set weights to NULL to weight the data uniformly by 1 (default).

Note that the weights array is *not* copied and must be valid througout the object lifetime (unless it is replaced).

Definition at line 1320 of file svm.c.

VL_EXPORT void vl_svm_train ( VlSvm * self )

Run the SVM solver.

Parameters:

self object.

The data on which the SVM operates is passed upon the cration of the VlSvm object. This function runs a solver to learn a corresponding model. See Getting started.

Definition at line 2156 of file svm.c.

Classes

Typedefs

Enumerations

Functions

Detailed Description

Typedef Documentation

Enumeration Type Documentation

Function Documentation