ML::SVM Namespace Reference

Classes
struct	LinearSVMParams
struct	LinearSVMModel
class	SVC
	C-Support Vector Classification. More...
struct	SupportStorage
struct	SvmModel
struct	SvmParameter

Enumerations
enum	SvmType { C_SVC , NU_SVC , EPSILON_SVR , NU_SVR }

Functions
template<typename math_t >
void	svcFit (const raft::handle_t &handle, math_t *input, int n_rows, int n_cols, math_t *labels, const SvmParameter &param, raft::distance::kernels::KernelParams &kernel_params, SvmModel< math_t > &model, const math_t *sample_weight)
	Fit a support vector classifier to the training data. More...
template<typename math_t >
void	svcFitSparse (const raft::handle_t &handle, int *indptr, int *indices, math_t *data, int n_rows, int n_cols, int nnz, math_t *labels, const SvmParameter &param, raft::distance::kernels::KernelParams &kernel_params, SvmModel< math_t > &model, const math_t *sample_weight)
	Fit a support vector classifier to the training data. More...
template<typename math_t >
void	svcPredict (const raft::handle_t &handle, math_t *input, int n_rows, int n_cols, raft::distance::kernels::KernelParams &kernel_params, const SvmModel< math_t > &model, math_t *preds, math_t buffer_size, bool predict_class)
	Predict classes or decision function value for samples in input. More...
template<typename math_t >
void	svcPredictSparse (const raft::handle_t &handle, int *indptr, int *indices, math_t *data, int n_rows, int n_cols, int nnz, raft::distance::kernels::KernelParams &kernel_params, const SvmModel< math_t > &model, math_t *preds, math_t buffer_size, bool predict_class)
	Predict classes or decision function value for samples in input. More...
template<typename math_t >
void	svmFreeBuffers (const raft::handle_t &handle, SvmModel< math_t > &m)
template<typename math_t >
void	svrFit (const raft::handle_t &handle, math_t *X, int n_rows, int n_cols, math_t *y, const SvmParameter &param, MLCommon::Matrix::KernelParams &kernel_params, SvmModel< math_t > &model, const math_t *sample_weight=nullptr)
	Fit a support vector regressor to the training data. More...
template<typename math_t >
void	svrFitSparse (const raft::handle_t &handle, int *indptr, int *indices, math_t *data, int n_rows, int n_cols, int nnz, math_t *y, const SvmParameter &param, raft::distance::kernels::KernelParams &kernel_params, SvmModel< math_t > &model, const math_t *sample_weight=nullptr)
	Fit a support vector regressor to the training data. More...

Enumeration Type Documentation

SvmType

enum ML::SVM::SvmType

Enumerator
C_SVC
NU_SVC
EPSILON_SVR
NU_SVR

Function Documentation

svcFit()

template<typename math_t >

void ML::SVM::svcFit	(	const raft::handle_t &	handle,
		math_t *	input,
		int	n_rows,
		int	n_cols,
		math_t *	labels,
		const SvmParameter &	param,
		raft::distance::kernels::KernelParams &	kernel_params,
		SvmModel< math_t > &	model,
		const math_t *	sample_weight
	)

Fit a support vector classifier to the training data.

Each row of the input data stores a feature vector. We use the SMO method to fit the SVM.

The output device buffers in model shall be unallocated on entry.

Template Parameters

math_t

floating point type

Parameters

[in]	handle	the cuML handle
[in]	input	device pointer for the input data in column major format. Size n_rows x n_cols.
[in]	n_rows	number of rows
[in]	n_cols	number of columns
[in]	labels	device pointer for the labels. Size [n_rows].
[in]	param	parameters for training
[in]	kernel_params	parameters for the kernel function
[out]	model	parameters of the trained model
[in]	sample_weight	optional sample weights, size [n_rows]

svcFitSparse()

template<typename math_t >

void ML::SVM::svcFitSparse	(	const raft::handle_t &	handle,
		int *	indptr,
		int *	indices,
		math_t *	data,
		int	n_rows,
		int	n_cols,
		int	nnz,
		math_t *	labels,
		const SvmParameter &	param,
		raft::distance::kernels::KernelParams &	kernel_params,
		SvmModel< math_t > &	model,
		const math_t *	sample_weight
	)

Fit a support vector classifier to the training data.

Each row of the input data stores a feature vector. We use the SMO method to fit the SVM.

The output device buffers in model shall be unallocated on entry.

Template Parameters

math_t

floating point type

Parameters

[in]	handle	the cuML handle
[in]	indptr	device pointer for CSR row positions. Size [n_rows + 1].
[in]	indices	device pointer for CSR column indices. Size [nnz].
[in]	data	device pointer for the CSR data. Size [nnz].
[in]	n_rows	number of rows
[in]	n_cols	number of columns
[in]	nnz	number of stored entries.
[in]	labels	device pointer for the labels. Size [n_rows].
[in]	param	parameters for training
[in]	kernel_params	parameters for the kernel function
[out]	model	parameters of the trained model
[in]	sample_weight	optional sample weights, size [n_rows]

svcPredict()

template<typename math_t >

void ML::SVM::svcPredict	(	const raft::handle_t &	handle,
		math_t *	input,
		int	n_rows,
		int	n_cols,
		raft::distance::kernels::KernelParams &	kernel_params,
		const SvmModel< math_t > &	model,
		math_t *	preds,
		math_t	buffer_size,
		bool	predict_class
	)

Predict classes or decision function value for samples in input.

We evaluate the decision function f(x_i). Depending on the parameter predict_class, we either return f(x_i) or the label corresponding to sign(f(x_i)).

The predictions are calculated according to the following formulas:

$$f(x_i) = \sum_{j=1}^n_support K(x_i, x_j) * dual_coefs[j] + b)$$

pred(x_i) = label[sign(f(x_i))], if predict_class==true, or pred(x_i) = f(x_i), if predict_class==false.

Template Parameters

math_t

floating point type

Parameters

	handle	the cuML handle
[in]	input	device pointer for the input data in column major format, size [n_rows x n_cols].
[in]	n_rows	number of rows (input vectors)
[in]	n_cols	number of columns (features)
[in]	kernel_params	parameters for the kernel function
[in]	model	SVM model parameters
[out]	preds	device pointer to store the predicted class labels. Size [n_rows]. Should be allocated on entry.
[in]	buffer_size	size of temporary buffer in MiB
[in]	predict_class	whether to predict class label (true), or just return the decision function value (false)

svcPredictSparse()

template<typename math_t >

void ML::SVM::svcPredictSparse	(	const raft::handle_t &	handle,
		int *	indptr,
		int *	indices,
		math_t *	data,
		int	n_rows,
		int	n_cols,
		int	nnz,
		raft::distance::kernels::KernelParams &	kernel_params,
		const SvmModel< math_t > &	model,
		math_t *	preds,
		math_t	buffer_size,
		bool	predict_class
	)

Predict classes or decision function value for samples in input.

We evaluate the decision function f(x_i). Depending on the parameter predict_class, we either return f(x_i) or the label corresponding to sign(f(x_i)).

The predictions are calculated according to the following formulas:

$$f(x_i) = \sum_{j=1}^n_support K(x_i, x_j) * dual_coefs[j] + b)$$

pred(x_i) = label[sign(f(x_i))], if predict_class==true, or pred(x_i) = f(x_i), if predict_class==falsee.

Template Parameters

math_t

floating point type

Parameters

	handle	the cuML handle
[in]	indptr	device pointer for CSR row positions. Size [n_rows + 1].
[in]	indices	device pointer for CSR column indices. Size [nnz].
[in]	data	device pointer for the CSR data. Size [nnz].
[in]	n_rows	number of rows
[in]	n_cols	number of columns
[in]	nnz	number of stored entries.
[in]	kernel_params	parameters for the kernel function
[in]	model	SVM model parameters
[out]	preds	device pointer to store the predicted class labels. Size [n_rows]. Should be allocated on entry.
[in]	buffer_size	size of temporary buffer in MiB
[in]	predict_class	whether to predict class label (true), or just return the decision function value (false)

svmFreeBuffers()

template<typename math_t >

void ML::SVM::svmFreeBuffers	(	const raft::handle_t &	handle,
		SvmModel< math_t > &	m
	)

Deallocate device buffers in the SvmModel struct.

Parameters

[in]	handle	cuML handle
[in,out]	m	SVM model parameters

svrFit()

template<typename math_t >

void ML::SVM::svrFit	(	const raft::handle_t &	handle,
		math_t *	X,
		int	n_rows,
		int	n_cols,
		math_t *	y,
		const SvmParameter &	param,
		MLCommon::Matrix::KernelParams &	kernel_params,
		SvmModel< math_t > &	model,
		const math_t *	sample_weight = nullptr
	)

Fit a support vector regressor to the training data.

Each row of the input data stores a feature vector.

The output buffers in model shall be unallocated on entry.

Template Parameters

math_t

floating point type

Parameters

[in]	handle	the cuML handle
[in]	X	device pointer for the input data in column major format. Size n_rows x n_cols.
[in]	n_rows	number of rows
[in]	n_cols	number of columns
[in]	y	device pointer for target values. Size [n_rows].
[in]	param	parameters for training
[in]	kernel_params	parameters for the kernel function
[out]	model	parameters of the trained model
[in]	sample_weight	optional sample weights, size [n_rows]

svrFitSparse()

template<typename math_t >

void ML::SVM::svrFitSparse	(	const raft::handle_t &	handle,
		int *	indptr,
		int *	indices,
		math_t *	data,
		int	n_rows,
		int	n_cols,
		int	nnz,
		math_t *	y,
		const SvmParameter &	param,
		raft::distance::kernels::KernelParams &	kernel_params,
		SvmModel< math_t > &	model,
		const math_t *	sample_weight = nullptr
	)

Fit a support vector regressor to the training data.

Each row of the input data stores a feature vector.

The output buffers in model shall be unallocated on entry.

Template Parameters

math_t

floating point type

Parameters

[in]	handle	the cuML handle
[in]	indptr	device pointer for CSR row positions. Size [n_rows + 1].
[in]	indices	device pointer for CSR column indices. Size [nnz].
[in]	data	device pointer for the CSR data. Size [nnz].
[in]	n_rows	number of rows
[in]	n_cols	number of columns
[in]	nnz	number of stored entries.
[in]	y	device pointer for target values. Size [n_rows].
[in]	param	parameters for training
[in]	kernel_params	parameters for the kernel function
[out]	model	parameters of the trained model
[in]	sample_weight	optional sample weights, size [n_rows]