A polynomial kernel function for a linear classifier to adapt it to data that is not linearly separable. More...

#include <polynomial.h>

Public Member Functions
	polynomial (uint8_t power=1, double c=1)
	Constructs a new polynomial kernel with the given parameters. More...

template<class PostingsData >
double	operator() (const PostingsData &first, const PostingsData &second) const
	Computes the value of \(K(first, second)\).

Private Attributes
uint8_t	power_
	\(p\), the power for the kernel

double	c_
	\(c\), the scalar term for the kernel

Detailed Description

A polynomial kernel function for a linear classifier to adapt it to data that is not linearly separable.

Uses the general form of: \(K(x,z) = (x^T z + c)^p\)

Constructor & Destructor Documentation

meta::classify::kernel::polynomial::polynomial	(	uint8_t	power = `1`,
		double	c = `1`
	)

inline

Constructs a new polynomial kernel with the given parameters.

Parameters

power	\(p\), the power for the polynomial kernel.
c	\(c\), the additional scalar term to allow for the use of existing features in the original space

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