SVM< TKernel > Class Template Reference

Class for SVM. More...

Collaboration diagram for SVM< TKernel >:

[legend]

List of all members.

Public Types
typedef TKernel	Kernel
Public Member Functions
void	BatchPredict (int learner_typeid, Dataset &testset, String predictedvalue_filename)
	Online batch classification for multiple testing vectors.
void	Init (int learner_typeid, const Dataset &dataset, datanode *module)
	SVM initialization.
void	InitTrain (int learner_typeid, const Dataset &dataset, datanode *module)
	Initialization(data dependent) and training for SVM learners.
void	LoadModelBatchPredict (int learner_typeid, Dataset &testset, String model_filename, String predictedvalue_filename)
	Load models from a file, and perform offline batch classification for multiple testing vectors.
double	Predict (int learner_typeid, const Vector &vector)
	SVM prediction for one testing vector.

---

Detailed Description

template<typename TKernel>
class SVM< TKernel >

Class for SVM.

Definition at line 119 of file svm.h.

---

Member Function Documentation

template<typename TKernel >

void SVM< TKernel >::BatchPredict	(	int	learner_typeid,
		Dataset &	testset,
		String	predictedvalue_filename
	)			[inline]

Online batch classification for multiple testing vectors.

No need to load model file, since models are already in RAM.

Note: for test set, if no true test labels provided, just put some dummy labels (e.g. all -1) in the last row of testset

Parameters:

		type id of the learner
		testing set
		file name of the testing data

Definition at line 629 of file svm.h.

References Dataset::matrix(), Dataset::n_features(), Dataset::n_points(), and SVM< TKernel >::Predict().

Referenced by SVM< TKernel >::LoadModelBatchPredict(), and main().

template<typename TKernel >

void SVM< TKernel >::Init	(	int	learner_typeid,
		const Dataset &	dataset,
		datanode *	module
	)			[inline]

SVM initialization.

Parameters:

		labeled training set or testing set
		number of classes (different labels) in the data set
		module name

Definition at line 220 of file svm.h.

References fx_param_double(), fx_param_int(), fx_submodule(), ArrayList< TElem >::Init(), Dataset::n_features(), Dataset::n_labels(), and Dataset::n_points().

Referenced by SVM< TKernel >::InitTrain(), and main().

template<typename TKernel >

void SVM< TKernel >::InitTrain	(	int	learner_typeid,
		const Dataset &	dataset,
		datanode *	module
	)			[inline]

Initialization(data dependent) and training for SVM learners.

Parameters:

		typeid of the learner
		number of classes (different labels) in the training set
		module name

Definition at line 286 of file svm.h.

References SVM< TKernel >::Init().

Referenced by main().

template<typename TKernel >

void SVM< TKernel >::LoadModelBatchPredict	(	int	learner_typeid,
		Dataset &	testset,
		String	model_filename,
		String	predictedvalue_filename
	)			[inline]

Load models from a file, and perform offline batch classification for multiple testing vectors.

Parameters:

		type id of the learner
		testing set
		name of the model file
		name of the file to store classified labels

Definition at line 662 of file svm.h.

References SVM< TKernel >::BatchPredict().

Referenced by main().

template<typename TKernel >

double SVM< TKernel >::Predict	(	int	learner_typeid,
		const Vector &	datum
	)			[inline]

SVM prediction for one testing vector.

Parameters:

		type id of the learner
		testing vector

Returns:

: predited value

Definition at line 511 of file svm.h.

Referenced by SVM< TKernel >::BatchPredict().

---

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

• mlpack/svm/svm.h