ml::SVM< T > Class Template Reference

#include <svm.hpp>

Public Member Functions
	SVM (KernelType kernel_type=KernelType::RBF, T C=1.0, T gamma=1.0, int degree=3, T coef0=0.0, T tol=1e-3, size_t max_iter=1000, T learning_rate=0.01)
	Constructor.
void	fit (const Matrix< T > &X, const std::vector< int > &y)
	Fit the SVM model to training data using automatic differentiation.
std::vector< int >	predict (const Matrix< T > &X) const
	Predict class labels for samples.
Matrix< T >	predict_proba (const Matrix< T > &X) const
	Predict class probabilities for samples.
std::vector< T >	decision_function (const Matrix< T > &X) const
	Compute the decision function for samples.
Matrix< T >	support_vectors () const
	Get support vectors.
std::vector< size_t >	support () const
	Get support vector indices.
std::vector< T >	dual_coef () const
	Get dual coefficients.
T	intercept () const
	Get intercept term.
std::vector< T >	loss_history () const
	Get training loss history.

Detailed Description

template<typename T>
class ml::SVM< T >

Definition at line 52 of file svm.hpp.

Constructor & Destructor Documentation

SVM()

template<typename T >

ml::SVM< T >::SVM	(	KernelType	kernel_type = KernelType::RBF,
		T	C = 1.0,
		T	gamma = 1.0,
		int	degree = 3,
		T	coef0 = 0.0,
		T	tol = 1e-3,
		size_t	max_iter = 1000,
		T	learning_rate = 0.01
	)

Constructor.

Parameters

kernel_type	Type of kernel function to use
C	Regularization parameter
gamma	Kernel coefficient for RBF, polynomial and sigmoid kernels
degree	Degree for polynomial kernel
coef0	Independent term for polynomial and sigmoid kernels
tol	Tolerance for stopping criterion
max_iter	Maximum number of iterations
learning_rate	Learning rate for gradient descent

Definition at line 12 of file svm.cpp.

Member Function Documentation

decision_function()

template<typename T >

std::vector< T > ml::SVM< T >::decision_function

(

const Matrix< T > &

X

)

const

Compute the decision function for samples.

Parameters

X	Input samples matrix (n_samples x n_features)

Returns

Vector of decision function values

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 290 of file svm.cpp.

dual_coef()

template<typename T >

std::vector< T > ml::SVM< T >::dual_coef

(

)

const

Get dual coefficients.

Returns

Vector of dual coefficients

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 342 of file svm.cpp.

fit()

template<typename T >

void ml::SVM< T >::fit	(	const Matrix< T > &	X,
		const std::vector< int > &	y
	)

Fit the SVM model to training data using automatic differentiation.

Parameters

X	Training features matrix (n_samples x n_features)
y	Training labels vector

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 31 of file svm.cpp.

intercept()

template<typename T >

T ml::SVM< T >::intercept

(

)

const

Get intercept term.

Returns

Intercept value

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 347 of file svm.cpp.

loss_history()

template<typename T >

std::vector< T > ml::SVM< T >::loss_history ( ) const

inline

Get training loss history.

Returns

Vector of loss values during training

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 124 of file svm.hpp.

predict()

template<typename T >

std::vector< int > ml::SVM< T >::predict

(

const Matrix< T > &

X

)

const

Predict class labels for samples.

Parameters

X	Input samples matrix (n_samples x n_features)

Returns

Vector of predicted class labels

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 274 of file svm.cpp.

predict_proba()

template<typename T >

Matrix< T > ml::SVM< T >::predict_proba

(

const Matrix< T > &

X

)

const

Predict class probabilities for samples.

Parameters

X	Input samples matrix (n_samples x n_features)

Returns

Matrix of class probabilities (n_samples x n_classes)

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 317 of file svm.cpp.

support()

template<typename T >

std::vector< size_t > ml::SVM< T >::support

(

)

const

Get support vector indices.

Returns

Vector of support vector indices

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 337 of file svm.cpp.

support_vectors()

template<typename T >

Matrix< T > ml::SVM< T >::support_vectors

(

)

const

Get support vectors.

Returns

Matrix of support vectors

Examples

/home/runner/work/deep-learning-algo-impls/deep-learning-algo-impls/include/ml/svm.hpp.

Definition at line 332 of file svm.cpp.

---

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

• include/ml/svm.hpp
• src/ml/svm.cpp