Public Member Functions | Protected Member Functions | Static Protected Member Functions

CKernelLocallyLinearEmbedding Class Reference


Detailed Description

the class KernelLocallyLinearEmbedding used to preprocess data using kernel extension of Locally Linear Embedding algorithm as described in

Decoste, D. (2001). Visualizing Mercer Kernel Feature Spaces Via Kernelized Locally-Linear Embeddings. The 8th International Conference on Neural Information Processing ICONIP2001

It is optimized with alignment formulation as described in

Zhao, D. (2006). Formulating LLE using alignment technique. Pattern Recognition, 39(11), 2233-2235. Retrieved from http://linkinghub.elsevier.com/retrieve/pii/S0031320306002160

Definition at line 42 of file KernelLocallyLinearEmbedding.h.

Inheritance diagram for CKernelLocallyLinearEmbedding:
Inheritance graph
[legend]

List of all members.

Public Member Functions

 CKernelLocallyLinearEmbedding ()
 CKernelLocallyLinearEmbedding (CKernel *kernel)
virtual ~CKernelLocallyLinearEmbedding ()
virtual CFeaturesapply (CFeatures *features)
CSimpleFeatures< float64_t > * embed_kernel (CKernel *kernel)
virtual const char * get_name () const

Protected Member Functions

virtual SGMatrix< float64_tconstruct_weight_matrix (SGMatrix< float64_t > kernel_matrix, SGMatrix< int32_t > neighborhood_matrix)
 HELPERS.
virtual SGMatrix< int32_t > get_neighborhood_matrix (SGMatrix< float64_t > kernel_matrix, int32_t k)

Static Protected Member Functions

static void * run_neighborhood_thread (void *p)
 THREADS.
static void * run_linearreconstruction_thread (void *p)

Constructor & Destructor Documentation

constructor

Definition at line 75 of file KernelLocallyLinearEmbedding.cpp.

constructor

Parameters:
kernel kernel to be used

Definition at line 80 of file KernelLocallyLinearEmbedding.cpp.

~CKernelLocallyLinearEmbedding (  )  [virtual]

destructor

Definition at line 91 of file KernelLocallyLinearEmbedding.cpp.


Member Function Documentation

CFeatures * apply ( CFeatures features  )  [virtual]

apply

Reimplemented from CLocallyLinearEmbedding.

Definition at line 95 of file KernelLocallyLinearEmbedding.cpp.

SGMatrix< float64_t > construct_weight_matrix ( SGMatrix< float64_t kernel_matrix,
SGMatrix< int32_t >  neighborhood_matrix 
) [protected, virtual]

HELPERS.

construct weight matrix

Reimplemented in CKernelLocalTangentSpaceAlignment.

Definition at line 129 of file KernelLocallyLinearEmbedding.cpp.

CSimpleFeatures< float64_t > * embed_kernel ( CKernel kernel  ) 

embed kernel (kernel should be inited)

Parameters:
kernel kernel to construct embed

Definition at line 115 of file KernelLocallyLinearEmbedding.cpp.

const char * get_name (  )  const [virtual]

get name

Reimplemented from CLocallyLinearEmbedding.

Reimplemented in CKernelLocalTangentSpaceAlignment.

Definition at line 86 of file KernelLocallyLinearEmbedding.cpp.

SGMatrix< int32_t > get_neighborhood_matrix ( SGMatrix< float64_t kernel_matrix,
int32_t  k 
) [protected, virtual]

construct neighborhood matrix by kernel matrix

Parameters:
kernel_matrix kernel matrix to be used
k k
Returns:
matrix containing indexes of neighbors of i-th object in i-th column

Reimplemented from CLocallyLinearEmbedding.

Definition at line 265 of file KernelLocallyLinearEmbedding.cpp.

void * run_linearreconstruction_thread ( void *  p  )  [static, protected]

runs linear reconstruction thread

Parameters:
p thread params

Reimplemented from CLocallyLinearEmbedding.

Definition at line 195 of file KernelLocallyLinearEmbedding.cpp.

void * run_neighborhood_thread ( void *  p  )  [static, protected]

THREADS.

runs neighborhood determination thread

Parameters:
p thread params

Reimplemented from CLocallyLinearEmbedding.

Definition at line 324 of file KernelLocallyLinearEmbedding.cpp.


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