hopfield离散matlab,神经网络系列讲解之离散型Hopfield网络

% ------------------------------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;-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;-1 1 1 1 1 1 1 1 1 -1;...

-1 -1 -1 -1 -1 -1 -1 -1 -1 -1];

%  --------------------------plot standard number figure--------------

subplot(2,3,1)

imshow(imresize(one,20))

title('standard number')

subplot(2,3,4)

imshow(imresize(two,20))

title('standard number')

% ---------------------------creat hopfield net---------------------

T=[one;two]';

net=newhop(T);

% --------------------------generate rand noise------------------

for i=2:11

for j=2:9

a=rand;

if a<=0.1

one(i,j)=-one(i,j);

two(i,j)=-two(i,j);

end

end

end

noise_one=one

noise_two=two

% -------------------------plot noise figure----------------------------

subplot(2,3,2)

imshow(imresize(noise_one,20))

title('noise number')

subplot(2,3,5)

imshow(imresize(noise_two,20))

title('noise number')

% ------------------------plot identify figure---------------------------

noise1={(noise_one)'};

tu1=sim(net,{12,3},{},noise1);

tu1{3}'

subplot(2,3,3)

imshow(imresize(tu1{3}',20))

title('identify number')

noise2={(noise_two)'};

tu2=sim(net,{12,3},{},noise2);

tu2{3}'

subplot(2,3,6)

imshow(imresize(tu2{3}',20))

title('identify number')