e_greedykpca.asv

具有图形操作界面的支持向量机多类分类实验系统.全部用Matlab实现,可以实现多种分类识别. 这是本人的毕业设计的附属程序

%addpath G:\Matlab_EXP\stprtool
%addpath G:\Matlab_EXP\stprtool\data%\riply_data


X = gencircledata([1;1],5,250,1); % generate training data
subplot(2,1,1);title('原始分布');
 ppatterns(X);
options.ker ='rbf'; % use RBF kernel
options.arg = 4; % kernel argument
options.new_dim = 2; % output dimension
options.m = 25; % size of sel. subset
model = greedykpca(X,options); % run greedy algorithm
XR = kpcarec(X,model); % reconstructed vectors
subplot(2,1,1);title('原始分布');
 ppatterns(orig_data);
h1 = ppatterns(X);pause;
h2 = ppatterns(XR,'+r');pause;
h3 = ppatterns(model.sv.X,'ob',12);pause;
legend([h1 h2 h3], ...
'Training set','Reconstructed set','Selected subset');