Public Member Functions | Protected Member Functions | Protected Attributes

CAUCKernel Class Reference


Detailed Description

The AUC kernel can be used to maximize the area under the receiver operator characteristic curve (AUC) instead of margin in SVM training.

It takes as argument a sub-kernel and Labels based on which number of positive labels times number of negative labels many ``virtual'' examples are created that ensure that all positive examples get a higher score than all negative examples in training.

Definition at line 33 of file AUCKernel.h.

Inheritance diagram for CAUCKernel:
Inheritance graph
[legend]

List of all members.

Public Member Functions

 CAUCKernel (void)
 CAUCKernel (int32_t size, CKernel *subkernel)
virtual ~CAUCKernel ()
CLabelssetup_auc_maximization (CLabels *labels)
virtual bool init (CFeatures *l, CFeatures *r)
virtual EKernelType get_kernel_type ()
virtual const char * get_name () const
virtual EFeatureClass get_feature_class ()
virtual EFeatureType get_feature_type ()

Protected Member Functions

virtual float64_t compute (int32_t idx_a, int32_t idx_b)

Protected Attributes

CKernelsubkernel

Constructor & Destructor Documentation

CAUCKernel ( void   ) 

default constructor

Definition at line 27 of file AUCKernel.cpp.

CAUCKernel ( int32_t  size,
CKernel subkernel 
)

constructor

Parameters:
size cache size
subkernel the subkernel

Definition at line 33 of file AUCKernel.cpp.

~CAUCKernel (  )  [virtual]

destructor

Definition at line 40 of file AUCKernel.cpp.


Member Function Documentation

float64_t compute ( int32_t  idx_a,
int32_t  idx_b 
) [protected, virtual]

compute kernel function for features a and b idx_{a,b} denote the index of the feature vectors in the corresponding feature object

Parameters:
idx_a index a
idx_b index b
Returns:
computed kernel function at indices a,b

Reimplemented from CDotKernel.

Definition at line 132 of file AUCKernel.cpp.

virtual EFeatureClass get_feature_class (  )  [virtual]

return feature class the kernel can deal with

Returns:
feature class SIMPLE

Reimplemented from CDotKernel.

Definition at line 83 of file AUCKernel.h.

virtual EFeatureType get_feature_type (  )  [virtual]

return feature type the kernel can deal with

Returns:
word feature type

Reimplemented from CDotKernel.

Definition at line 89 of file AUCKernel.h.

virtual EKernelType get_kernel_type (  )  [virtual]

return what type of kernel we are

Returns:
kernel type AUC

Implements CDotKernel.

Definition at line 71 of file AUCKernel.h.

virtual const char* get_name ( void   )  const [virtual]

return the kernel's name

Returns:
name AUC

Reimplemented from CDotKernel.

Definition at line 77 of file AUCKernel.h.

bool init ( CFeatures l,
CFeatures r 
) [virtual]

initialize kernel

Parameters:
l features of left-hand side
r features of right-hand side
Returns:
if initializing was successful

Reimplemented from CDotKernel.

Definition at line 125 of file AUCKernel.cpp.

CLabels * setup_auc_maximization ( CLabels labels  ) 

initialize kernel based on current labeling and subkernel

Parameters:
labels - current labeling
Returns:
new label object to be used together with this kernel in SVM training for AUC maximization

Definition at line 46 of file AUCKernel.cpp.


Member Data Documentation

CKernel* subkernel [protected]

the subkernel

Definition at line 104 of file AUCKernel.h.


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines

SHOGUN Machine Learning Toolbox - Documentation