How to form my target vector for neural network to classify eeg signal?

I am trying to classify eeg signals using neural network from the data collected form 5 tasks(classes). I obtain my data from 7 electrode. The data size for each task is 7x2500. I applied a notch and bandpass filter to eliminate noise. I then performed a 64-point fft to individual tasks row wise to give a matrix of 7x64. From this fft matrix I made it into a single vector, that is each column was transposed into rows to give 448x1 vector. 5 such vectors were made for 5 tasks and concatenated into a single matrix to give a 448x5 matrix. This will be my input matrix in the neural network. I m stuck in the target vector. I have 5 classes. Pls help me with the target vector. Thanks.

Answer 1

[Greg Heath]

The target matrix for classifying N I-dimensional input vectors into c classes is a c X N matrix containing columns of the eye(c) unit matrix. The conversion from true class indices to a target matrix is obtained via the function IND2VEC

The estimated class indices are obtained from the output matrix via the function VEC2IND. Errors are obtained by comparing estimated indices with the true indices.

 help patternnet
 doc patternnet

Search both the NEWSGROUP and ANSWERS using

 patternnet greg

Hope this helps.

Thank you for formally accepting my answer

Greg