How can I use Six different target matrices and One input matrix on Neural Network?

I have six different matrices resulting from training of six classes, with each having O's and 1's. Please how do I combine it so I could produce a single confusion matrix using the Neural Network Toolbox. My input matrix already has in it 1,2,3,4,5,6 to show six different classes already gotten. Thank you in advance.

Answer 1

[Greg Heath]

 [ I N ] = size(input)% N I-dimensional input vectors

 [c N ] = size(target)% corresponding N c-dimensional {0,1} unit target vectors for c=6 classes

% sum(target) = ones(1,N)

 trueclassindices = vec2ind(target)

 target = ind2vec(trueclassindices)

Hope this helps

Thank you for formally accepting my answer

Greg

[bayoishola20]

Sorry I don't really get this. Can't I do this on NNToolbox? Thanks in anticipation. N:B I'm just fairly good in using MATLAB.

[Greg Heath]

I am confused.

What exactly is your problem?

%I have six different matrices resulting from training of six classes, with each having O's and 1's.

Not clear why you have 6 matrices after training. Please explain. Are these output matrices? What size is each matrix.? Are these training, validation or testing results?

>Please how do I combine it so I could produce a single confusion matrix using the Neural Network Toolbox. My input matrix already has in it 1,2,3,4,5,6 to show six different classes already gotten. Thank you in advance.

 Your input matrix??? To clarify, explain the size and structure of input, target and output matrices.

[bayoishola20]

I got the classes trained using the

roitool

and I have each with 0's and 1's [the size of each is 2408*4628]. But my segmented images has 1,2,3,4,5,6 classes [2408*4628]. I need to at the end of the day get a confusion matrix using the NNToolbox. Thanks in anticipation.

[Greg Heath]

>> lookfor roitool roitool not found.

Sorry, I am still confused.

Perhaps IMAGEANALYST or someone else can help

Greg

[bayoishola20]

sorry, I meant roipoly.