Data training with backpropagation

Subject:	Data training with backpropagation
Date:	Mon, 8 Feb 2010 23:53:07 -0800 (PST)

Hi,
i would like to train network using backpropagation in matlab with
large data where the network structure is 5 nodes for input 6 nodes
hideen 4 nodes output.
i should recognize data of four class ( A ,B,C,D ) some of our data
are as follows:
(A)
(B) (C)
(D) 1 1
1 1 1 1
1 1
-0.00281 -0.00276 -0.00248 0.019727 -0.000696
-0.002768 -0.001826 -0.001998
-0.09323 -0.09280 -0.09280 -0.090197 -0.093245
-0.093217 -0.093170 -0.093217
-0.99987 -1.00005 -1.00002 -0.992684 -1.000050
-1.000230 -1.000199 -1.000183
-0.15744 -0.15762 -0.15762 -0.156964 -0.157645
-0.157635 -0.157564 -0.157613

the target is as follows
t=[ 1 -1 -1 -1 ;
-1 1 -1 -1 ;
-1 -1 1 -1 ;
-1 -1 -1 1 ];

[net, tr]= train( net, pn, t);

Should i train the first colum of data for all classes(A,B,C,D) and
when i get the same target return to the second colum of data
(A,B,C,D) ?
if you know any good example of training network with large data, may
you give me the link?
Thank you in advance.
Regards