FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR > Class Template Reference

Wraps external Support Vector Machine implementation (in LibSVM) to serve as FST3 classifier. More...

#include <classifier_svm.hpp>

Inheritance diagram for FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >:

[legend]

Collaboration diagram for FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >:

[legend]

List of all members.

Classes
class	ParameterSet
	Nested class to hold parameter candidates in the course of optimize_parameters() run. More...
Public Types
typedef boost::shared_ptr < DATAACCESSOR >	PDataAccessor
typedef boost::shared_ptr < SUBSET > const	PSubset
typedef DATAACCESSOR::PPattern	PPattern
Public Member Functions
void	initialize (const PDataAccessor da)
void	set_parameter_C (double newC)
void	set_parameter_gamma (double newgamma)
void	set_parameter_coef0 (double newcoef0)
void	set_kernel_type (int kernel_type)
double	get_parameter_C () const
double	get_parameter_gamma () const
double	get_parameter_coef0 () const
int	get_kernel_type () const
virtual bool	classify (DIMTYPE &cls, const PPattern &pattern)
	classifies pattern, returns the respective class index
virtual bool	train (const PDataAccessor da, const PSubset sub)
	learns from designated training part of data
virtual bool	test (RETURNTYPE &result, const PDataAccessor da)
	estimates accuracy using designated test data
bool	optimize_parameters (const PDataAccessor da, const PSubset sub, const int max_points=100, const int max_throws=100, const double lgC_min=-5, const double lgC_max=9, const double lggamma_min=-15, const double lggamma_max=3, const double lgcoef0_min=-2, const double lgcoef0_max=5, std::ostream &os=std::cout)
Classifier_LIBSVM *	clone () const
Classifier_LIBSVM *	sharing_clone () const
Classifier_LIBSVM *	stateless_clone () const
virtual std::ostream &	print (std::ostream &os) const
Protected Types
typedef list< ParameterSet >	PARAMSETLIST
Protected Member Functions
void	allocate ()
void	cleanup ()
Protected Attributes
IDXTYPE	_all_patterns
DIMTYPE	_classes
DIMTYPE	_features
struct svm_problem	problem
struct svm_parameter	parameters
struct svm_model *	model
struct svm_node *	onepattern
bool	svm_class_weighing
PARAMSETLIST::iterator	iter
Private Member Functions
	Classifier_LIBSVM (const Classifier_LIBSVM &csvm, int)
Private Attributes
boost::scoped_array< DIMTYPE >	_index
DIMTYPE	_subfeatures

---

Detailed Description

template<class RETURNTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>
class FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >

Wraps external Support Vector Machine implementation (in LibSVM) to serve as FST3 classifier.

Note:

clone() implementation assumes LibSVM's svm_model and svm_node is equivalent to that in LibSVM version 300

Warning:

LIBSVM especially with LINEAR kernel seems to have occassional problems with certain C values on certain datasets and may freeze. (Other kernels are more stable but not completely immune to this problem.) This is a problem outside FST3.

Examples:

demo12t.cpp, demo23.cpp, demo25t.cpp, demo32t.cpp, demo35t.cpp, demo53.cpp, and demo62.cpp.

---

Member Function Documentation

template<class RETURNTYPE , typename IDXTYPE , typename DIMTYPE , class SUBSET , class DATAACCESSOR >

bool FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::optimize_parameters	(	const PDataAccessor	da,
		const PSubset	sub,
		const int	max_points = 100,
		const int	max_throws = 100,
		const double	lgC_min = -5,
		const double	lgC_max = 9,
		const double	lggamma_min = -15,
		const double	lggamma_max = 3,
		const double	lgcoef0_min = -2,
		const double	lgcoef0_max = 5,
		std::ostream &	os = std::cout
	)			[inline]

Warning:

LIBSVM LINEAR kernel seems to have problems with certain C values on certain datasets. If the optimization process seemingly freezes, try narrower lgC_min and lgC_max. Especially the lower bound seems important to be increased.

References FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::test(), and FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::train().

template<class RETURNTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>

Classifier_LIBSVM* FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::clone

(

)

const [inline, virtual]

create 1:1 independent clone of the current object

Implements FST::Clonable.

References FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::stateless_clone().

Referenced by FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::stateless_clone().

template<class RETURNTYPE, typename IDXTYPE, typename DIMTYPE, class SUBSET, class DATAACCESSOR>

Classifier_LIBSVM* FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::sharing_clone

(

)

const [inline, virtual]

Note:

dirty workaround to enable easy usage in sequential_step ...

Implements FST::Clonable.

template<class RETURNTYPE , typename IDXTYPE , typename DIMTYPE , class SUBSET , class DATAACCESSOR >

Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR > * FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::stateless_clone

(

)

const [inline, virtual]

create clone of the current object, ignoring internal temporary structures to save speed. Does not replicate exact object state. The clone must be used carefully in a way that ensures internal structure re-initialization Use example: Data_Splitter cloning or Classifier_SVM cloning due to inability to clone external structures defined in LibSVM

Implements FST::Clonable.

References FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::clone().

Referenced by FST::Classifier_LIBSVM< RETURNTYPE, IDXTYPE, DIMTYPE, SUBSET, DATAACCESSOR >::clone().

---

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

• classifier_svm.hpp