need help with neural network

1 view (last 30 days)
amro mahran
amro mahran on 23 Jan 2012
i'm working on a M file (script) to run neural network training for several time (10 times each) and i want to save the neural network results concerning MSE and Regression R coefficient for the Training and Validation and test data i want to know how to write a code to save the neural results??
  1 Comment
amro mahran
amro mahran on 24 Jan 2012
I want to know the code of the function that calculate the MSE and Rcoefficient

Sign in to comment.

Accepted Answer

Léon
Léon on 24 Jan 2012
The solution to your problem should be a simple loop. Meaning that you put your relevant ANN code in that loop and save the output each time. Like the following
MSE_all = zeros(10,1); % preallocate space to the vector for all MSE values
R_all = zeros(10,1); % see above
for i = 10
% here goes your ANN training code
MSE_all(i) = MSE; % MSE means here the needed output by the ANN that should
% be stored
R_all(i) = R; % see above
end
  1 Comment
amro mahran
amro mahran on 24 Jan 2012
Thanks for your help
but please i want to know the MSE function and the R function that is used

Sign in to comment.

More Answers (2)

Greg Heath
Greg Heath on 25 Jan 2012
Many examples of code that computes and saves Ntrials values of MSE and R^2 (coefficient of determination) can be obtained by searching the Newsgroup with the search words
heath Ntrials
Hope this helps
Greg
  1 Comment
amro mahran
amro mahran on 25 Jan 2012
I tried to understand Newsgroup results but i' m a beginner
i'm using the Plotreg() function to determine the Regression R coefficient as follow
Y=sim(net,inputs0);
[~,~,r]=postreg(Y(tr.testInd),targets(tr.testInd))
is it right or not????
and for the MSE i use:
E=targets-Y;
mse_test=mse(E(tr.testInd));
Is it wright ???

Sign in to comment.


Greg Heath
Greg Heath on 29 Jan 2012
If you have source access try this:
type mse
type postreg
Hope this helps.
Greg

Categories

Find more on Deep Learning Toolbox in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!