离散Hopfield神经网络
原文链接：https://blog.csdn.net/linuxcumt1/article/details/8597786
-
% ------------------------standard number
array-----------------------
-
-
one=[
-1
-1
-1
-1
1
1
-1
-1
-1
-1;
-1
-1
-1
-1
1
1
-1
-1
-1
-1;...
-
-1
-1
-1
-1
1
1
-1
-1
-1
-1;
-1
-1
-1
-1
1
1
-1
-1
-1
-1;...
-
-1
-1
-1
-1
1
1
-1
-1
-1
-1;
-1
-1
-1
-1
1
1
-1
-1
-1
-1;...
-
-1
-1
-1
-1
1
1
-1
-1
-1
-1;
-1
-1
-1
-1
1
1
-1
-1
-1
-1;...
-
-1
-1
-1
-1
1
1
-1
-1
-1
-1;
-1
-1
-1
-1
1
1
-1
-1
-1
-1];
-
-
two=[
-1
1
1
1
1
1
1
1
1
-1;
-1
1
1
1
1
1
1
1
1
-1;...
-
-1
-1
-1
-1
-1
-1
-1
1
1
-1;
-1
-1
-1
-1
-1
-1
-1
1
1
-1;...
-
-1
1
1
1
1
1
1
1
1
-1;
-1
1
1
1
1
1
1
1
1
-1;...
-
-1
1
1
-1
-1
-1
-1
-1
-1
-1;
-1
1
1
-1
-1
-1
-1
-1
-1
-1;...
-
-1
1
1
1
1
1
1
1
1
-1;
-1
1
1
1
1
1
1
1
1
-1];
-
-
% ----------------------plot standard number figure-----------------
-
-
ONE=imresize(one,
20);
-
subplot(
3,
2,
1)
-
imshow(ONE)
-
title(
'stand number')
-
-
TWO=imresize(two,
20);
-
subplot(
3,
2,
2)
-
imshow(TWO)
-
title(
'stand number')
-
-
% ----------------------create hopfield net--------------------------
-
-
T=[one;two]
';
-
net=newhop(T);
-
-
% ------------------------------generate noise----------------------------
-
% -------------------------noise array(fixed noise)--------------
-
-
% no1=[-1 -1 -1 -1 1 1 -1 -1 -1 -1;-1 -1 -1 -1 1 -1 -1 -1 -1 -1;...
-
% -1 -1 1 -1 1 1 -1 -1 -1 -1; -1 -1 -1 -1 1 1 1 -1 -1 -1;...
-
% -1 -1 -1 -1 -1 1 -1 -1 -1 -1; -1 -1 -1 -1 1 1 -1 1 -1 -1;...
-
% -1 -1 -1 -1 1 1 -1 -1 -1 -1; -1 1 -1 -1 1 1 -1 -1 -1 -1;...
-
% -1 -1 -1 -1 -1 1 -1 -1 -1 -1; -1 -1 -1 -1 1 1 -1 -1 1 -1];
-
%
-
% no2=[-1 1 1 1 -1 1 1 -1 1 -1;-1 1 1 1 1 1 1 1 1 -1;...
-
% -1 -1 1 -1 1 -1 -1 1 1 -1;-1 -1 -1 1 -1 1 -1 1 1 -1;...
-
% -1 1 1 1 1 1 1 1 1 -1;-1 1 1 1 1 1 1 1 1 -1;...
-
% -1 1 1 -1 -1 -1 -1 -1 -1 -1;-1 1 1 -1 -1 -1 -1 -1 -1 -1;...
-
% -1 1 1 1 -1 1 1 1 1 -1;-1 1 1 -1 1 1 1 1 1 -1];
-
-
% -----------------------noise array(rand noise)-----------------
-
-
rand('state',
0);
-
for i=
1:
100
-
a=rand;
-
if a<
0.1
-
one(i)=-one(i);
-
two(i)=-two(i);
-
end
-
end
-
no1=one;
-
no2=two;
-
-
% -------------------------plot noisy figure---------------
-
-
subplot(
3,
2,
3)
-
NO1=imresize(no1,
20);
-
imshow(NO1)
-
title(
'noisy number')
-
-
subplot(
3,
2,
4)
-
NO2=imresize(no2,
20);
-
imshow(NO2)
-
title(
'noisy number')
-
-
% --------------------------plot identify figure--------------
-
-
noise1={(no1)
'};
-
tu1=sim(net,{10,50},{},noise1);
-
tu1{50}'
-
subplot(
3,
2,
5)
-
imshow(imresize(tu1{
50}
',20))
-
title('identify number')
-
-
noise2={(no2)
'};
-
tu2=sim(net,{10,50},{},noise2);
-
tu2{50}'
-
subplot(
3,
2,
6)
-
imshow(imresize(tu2{
50}
',20))
-
title('identify number')
-
-
web -browser http:
//www.ilovematlab.cn/thread-30439-1-1.html