SVM

Support Vector Machine's functions. More...

Functions
SIMD_API void	SimdSvmSumLinear (const float *x, const float *svs, const float *weights, size_t length, size_t count, float *sum)
	It is a part of linear SVM (Support Vector Machine) prediction algorithm. More...

Detailed Description

Support Vector Machine's functions.

Function Documentation

SimdSvmSumLinear()

void SimdSvmSumLinear	(	const float *	x,
		const float *	svs,
		const float *	weights,
		size_t	length,
		size_t	count,
		float *	sum
	)

It is a part of linear SVM (Support Vector Machine) prediction algorithm.

Algorithm's details:

sum = 0;
for(i = 0; i < count; ++i)
    for(j = 0; j < length; ++j)
        sum += x[j]*svs[j][i]*weight[i];

Note

The array with support vectors must has following structure: svs[length][count].

Parameters

[in]	x	- a vector of features which need to predict with using SVM.
[in]	svs	- an array with support vectors.
[in]	weights	- a weight coefficient of each support vector.
[in]	length	- a length of these current and support vectors.
[in]	count	- a count of support vectors.
[out]	sum	- a pointer to result sum.