Matlab代码编写的semi-supervised CCA 程序

clc
clear all 
close all
load （‘D:\Xin-E\database\SSLset=1data.mat‘）;
load （‘D:\Xin-E\database\SSLset=1splitslabeled=10.mat‘）;
train_num = 750;
X = X.‘;
P = size（X2）;
m_database = mean（X2）; 
A = X - repmat（m_database1P）; 
L = A‘*A;
[VD] = eig（L）;
[VD] = sortem（VD）;
Y = X*V;

k = 20;
delta_0 = 0.5;