C++ Programming
02-20-2010, 03:00 AM
hi..
i have asked question about neural network before.thank you for those who replied..
actually, i'm working with mlp backpropagation using matlab. i have about 20o continuous value inputs with 200 continuous outputs. each input has different value of output. i don't have any idea to write the full matlab code for the training. For classification problem, as example,i can use this code to convert the output to 0 or 1:
%convert outtrainset
for j=1:2
for i=1:2000;
if outtrainset(i,j)>=0.5;
outtrainconvert(i,j)=1;
else
outtrainconvert(i,j)=0;
end
end
end
What should i do in modify this code, so that the outputs are continuous values, not 0 or 1. i really need help. hope someone can help me and give me example of related code.tq for any reply..
i have asked question about neural network before.thank you for those who replied..
actually, i'm working with mlp backpropagation using matlab. i have about 20o continuous value inputs with 200 continuous outputs. each input has different value of output. i don't have any idea to write the full matlab code for the training. For classification problem, as example,i can use this code to convert the output to 0 or 1:
%convert outtrainset
for j=1:2
for i=1:2000;
if outtrainset(i,j)>=0.5;
outtrainconvert(i,j)=1;
else
outtrainconvert(i,j)=0;
end
end
end
What should i do in modify this code, so that the outputs are continuous values, not 0 or 1. i really need help. hope someone can help me and give me example of related code.tq for any reply..