CDomainAdaptationSVM Class Reference


Detailed Description

class DomainAdaptiveSVM

Definition at line 23 of file DomainAdaptationSVM.h.

List of all members.

Public Member Functions

 CDomainAdaptationSVM ()
 CDomainAdaptationSVM (float64_t C, CKernel *k, CLabels *lab, CSVM *presvm, float64_t B)
virtual ~CDomainAdaptationSVM ()
void init (CSVM *presvm, float64_t B)
virtual bool train (CFeatures *data=NULL)
virtual EClassifierType get_classifier_type ()
virtual CLabelsclassify (CFeatures *data)
virtual CSVMget_presvm ()
virtual float64_t get_B ()
virtual const char * get_name () const

Protected Member Functions

virtual bool is_presvm_sane ()

Protected Attributes

CSVMpresvm
float64_t B
float64_t train_factor

Constructor & Destructor Documentation

default constructor

CDomainAdaptationSVM ( float64_t  C,
CKernel k,
CLabels lab,
CSVM presvm,
float64_t  B 
)

constructor

Parameters:
C cost constant C
k kernel
lab labels
presvm trained SVM to regularize against
B trade-off constant B
virtual ~CDomainAdaptationSVM (  )  [virtual]

destructor


Member Function Documentation

virtual CLabels* classify ( CFeatures data  )  [virtual]

classify objects

Parameters:
data (test)data to be classified
Returns:
classified labels
virtual float64_t get_B (  )  [virtual]

getter for regularization parameter B

Returns:
regularization parameter B
virtual EClassifierType get_classifier_type (  )  [virtual]

get classifier type

Returns:
classifier type LIGHT

Definition at line 70 of file DomainAdaptationSVM.h.

virtual const char* get_name (  )  const [virtual]
Returns:
object name

Definition at line 96 of file DomainAdaptationSVM.h.

virtual CSVM* get_presvm (  )  [virtual]

returns SVM that is used as prior information

Returns:
presvm
void init ( CSVM presvm,
float64_t  B 
)

init SVM

Parameters:
presvm trained SVM to regularize against
B trade-off constant B
virtual bool is_presvm_sane (  )  [protected, virtual]

check sanity of presvm

Returns:
true if sane, throws SG_ERROR otherwise
virtual bool train ( CFeatures data = NULL  )  [virtual]

train SVM classifier

Parameters:
data training data (parameter can be avoided if distance or kernel-based classifiers are used and distance/kernels are initialized with train data)
Returns:
whether training was successful

Member Data Documentation

float64_t B [protected]

regularization parameter B

Definition at line 141 of file DomainAdaptationSVM.h.

CSVM* presvm [protected]

SVM to regularize against

Definition at line 137 of file DomainAdaptationSVM.h.

float64_t train_factor [protected]

flag to switch off regularization in training

Definition at line 145 of file DomainAdaptationSVM.h.


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

SHOGUN Machine Learning Toolbox - Documentation