Prev: dicominfo not returning RadionuclideTotalDose header field
Next: statistical t test to compare two groups
From: srikanthkurni on 11 May 2010 12:57 hi ppl.... i hv the following code in matlab p=[Vf1max;Vf2max;Vdmax;Vf1min;Vf2min;Vdmin;Vf1avg;Vf2avg;Vdavg]; t=[0 1]; net=newff(p,t,3); [net,tr]=train(net,p,t); a = sim(net,p); i got an error as follows: ??? Error using ==> network.train at 146 Targets are incorrectly sized for network. Matrix must have 1 columns. Error in ==> test at 161 [net,tr]=train(net,p,t); someone plz help me regarding this...
From: ade77 on 11 May 2010 18:21 srikanthkurni <srikanthkurni(a)gmail.com> wrote in message <564766261.118769.1273611477776.JavaMail.root(a)gallium.mathforum.org>... > hi ppl.... > i hv the following code in matlab > > p=[Vf1max;Vf2max;Vdmax;Vf1min;Vf2min;Vdmin;Vf1avg;Vf2avg;Vdavg]; > t=[0 1]; > net=newff(p,t,3); > [net,tr]=train(net,p,t); > a = sim(net,p); > > i got an error as follows: > > ??? Error using ==> network.train at 146 > Targets are incorrectly sized for network. > Matrix must have 1 columns. > > Error in ==> test at 161 > [net,tr]=train(net,p,t); > > someone plz help me regarding this... What is the size of your input?. You have output [0 1], yet you have 3 sets of input. Or maybe something like this: input = a1 a2 a3 b1 b2 b3 c1 c2 c3 output = 0 1 1 1 0 0 your first observation is (a1 b1 c1), and your target is (0 1). if you can clarify...
From: srikanthkurni on 12 May 2010 00:43 the inputs to neural network are 9 values defined in 'p'....
From: David Young on 12 May 2010 05:06
It may be that the problem is this: > t=[0 1]; which has 2 columns. The error message says: > Targets are incorrectly sized for network. > Matrix must have 1 columns. So your target matrix has 2 columns and the message says it must have 1 column. Check the documentation to see what the correct dimensions for the target are. It may be that t = [0; 1] would be OK (this has one column) but it may not be so simple. |