NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

Post by PROGRAMME » Thu, 05 May 2005 08:35:07


MATLAB USING NEURAL NETWORK TOOLBOX
this is my program!!! i am getting the error mentioned at the end.
x = -3.1416:0.25:3.1416; y = -3.1416:0.25:3.1416;
xlabel('xaxis');ylabel('yaxis');zlabel('zaxis');
title('surface Z2=e(sinx) e(-cos2y)');
P=[x;y]; T=z;
'purelin'},'trainlm');
%Define parameters
net.trainParam.show = 50;
net.trainParam.lr = 0.05;
net.trainParam.epochs = 300;
net.trainParam.goal = 1e-2;
%Train network
net1 = train(net, P, T);
gi=input('Strike any key ...');
ERROR:
??? Error using ==> network.train
Targets are incorrectly sized for network.
Matrix must have 17 rows.
KINDLY HELP ME ASAP!!!!
 
 
 

NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

Post by Duane Hans » Thu, 05 May 2005 08:44:35


Is your homework due by the end of this week?

Type everything in capital letters and you will get a different
response.

Sorry I don't know anything about neurotic networks.

 
 
 

NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

Post by Greg Heat » Thu, 05 May 2005 13:52:22


sizex = size(x) % x and y are row vectors


sizez = size(z) % z is a square matrix (which is not what you want)


This specifies 2 inputs, 25 hidden nodes and 17 outputs
which is not what you want.

Hope this helps.

Greg

sizez is not
 
 
 

NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

Post by PROGRAMME » Fri, 06 May 2005 01:26:48

thanks greg for replying ..BUT I COULD NOT DECIPHER WHAT U INTENDED
TO SAY...
I TRIED REDUCING THE HIDDEN NODES AND OUTPUTS(HAD DONE THAT BEFORE
TOO) FROM [25 17] TO [10 1] ETC BUT that did nto help.
i am not able to find whati s the real problem to this????
when i tried doing this using the range -2:0.25:2 for x and y the
whole thingworked perfectly but wheni am using it for -3.1416 aka pi
it is showing error.
KINDLY HELP !!
AWAITING UR REPLY.........
 
 
 

NEURAL NETWORK TOOLBOX ERROR.PLZ HELP ASAP

Post by PROGRAMME » Fri, 06 May 2005 01:46:55

FINALLY GOT IT!!!
THANK YOU VERY VERY MUCH!!
USEFUL KNOWLEDGE...BASICS ARE DAMN IMP!!!
minor errors...oh no!!!




the
you