erpanechnikovkernel.m

It is for Face Recognition

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% function erpanechnikovKernel()
%   z.li, 12-26-2003.
%   kernel regression on relevance function
% function dependency:
%   - n/a
% input:
%   n - power, 2, 3, etc
%   x0 - sample with known function value f(x0)
%   x1 - sample with un-known function value f(x1)
% output:
%   d - distance between x0, x1
%   p - prob in [0, 1]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%function [d, p]=erpanechnikovKernel(n, x0, x1)
function [d, p]=erpanechnikovKernel(n, x0, x1)
  d = norm(x0-x1, 2);
  if d > 1
    p = 0;
  else
    p = (1-d^n)^n;
  end