Package mystic :: Module svmtools

Module svmtools

source code

Simple utility functions for SV-Regressions

Functions
 
KernelMatrix(X, k) source code
 
SupportVectors(alpha, eps=0) source code
 
StandardInnerProduct(i1, i2) source code
 
LinearKernel(i1, i2) source code
 
Bias(x, y, alpha, eps, kernel=StandardInnerProduct)
Compute regression bias for epsilon insensitive loss regression
source code
 
RegressionFunction(x, y, alpha, eps, kernel=StandardInnerProduct)
The Support Vector expansion.
source code
Function Details

RegressionFunction(x, y, alpha, eps, kernel=StandardInnerProduct)

source code 

The Support Vector expansion. f(x) = Sum (ap - am) K(xi, x) + b