plotclassifiers.m

一个包含丰富内容的流形学习算法工具包

function plotclassifiers(classifier,xrange, yrange);

x=xrange;
y=yrange;

%x=-1.5:0.05:2.5; % x range
%y=-1.5:0.05:2.5; % y range

%x=0:0.01:1; % x range
%y=0:0.01:1; % y range


KERNELTYPE={'linear','poly','rbf'};

alpha=classifier.alpha;
Xtrain=classifier.xtrain;
b=classifier.b;
method=classifier.Name;
kerneltype=classifier.Kernel;
kernelparam=classifier.KernelParam;
lambda=classifier.gammas;


[xx,yy]=meshgrid(x,y); % makes a grid of points

X=[xx(:) yy(:)];


if (strcmp(method,'linear_rlsc') | strcmp(method,'linear_laprlsc'))

    z=X*alpha + b;

else

    K=calckernel(kerneltype,kernelparam, Xtrain, X);
 
    z=sign(K*alpha-b);
 
end

Z=reshape(z,length(x),length(y));


[cs,h]=contourf(xx,yy,Z,[0 0]);shading flat;
%clabel(cs,h);colorbar

hold on;
title([method '  ('  kerneltype  ') : ' '\sigma =' num2str(kernelparam) ]);
if size(lambda,2)==2
  xlabel([' \gamma_A = ' num2str(lambda(1)) '  \gamma_I = ' num2str(lambda(2))]);
else
  xlabel([' \gamma_I = ' num2str(lambda(1))]);  
end
%xlabel('White (-1) Green (+1)');
%title( [' \lambda_1 = ' num2str(lambda(1)) '   \lambda_2 = '  num2str(lambda(2))]);

%xlabel([' \lambda_1 = ' num2str(lambda(1)) ]);
%ylabel(['\lambda_2 = '  num2str(lambda(2))]);
 
%plot2D(Xtest,Ytest,5);