About neural network : newbie question

1 view (last 30 days)
I am a new to matlab. I just covert my image to matrix and want to train it. here is my coding:
P= [];
for i =1:15
str=strcat('E:Skripsi\crop2\image',int2str(i),'.bmp');
eval('img=imread(str);');
img=reshape(img,[],1);
P= [P img];
end
T=[1 1 1 1 1 1 1 1 1 1 0 0 0 0 0];// if true image i make it 1 , negative image make it 0
p = num2cell(P,1);
t = num2cell(T,1);
net = newff(minmax(p), [3 1], {'logsig' 'purelin'});
net.trainParam.epochs = 10;
net = train(net,p,t);
Y=sim(net,p)
these is what i got when i compile it
[0.9794] [0.9974] [1.2784] [1.0014] [1.0058] [0.9904] [0.7293] [1.0001] [1.0106]
Columns 10 through 15
[1.0002] [0.0618] [3.3131e-04] [9.9559e-04] [0.0134] [0.0178]
then my problem is what should i do to make the result success i train in percentage with error below 1% (what i want: example i train 10 true image to 1 and one result of my train is 0.798 then accurate =90%) thanks for the help.

Accepted Answer

Greg Heath
Greg Heath on 19 Dec 2013
There is no reason to use cells or only train for 10 epochs.
Use round to convert the output to {0,1}
Thank you for formally accepting my answer
Greg
  5 Comments
Parkz Leingodz
Parkz Leingodz on 20 Dec 2013
size P is 16x21 pixels.
P is a matrix that save whole matrix value of img.

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!