Plot a graph MSE vs number neurons in hidden layer on matlab

2 views (last 30 days)
I have a net with 7 inputs and 1 output.I want to do a study of convergence of the neurones in the hidden layer. So for that I need to do Plot a graph MSE vs number neurons in hidden layer on matlab. Anyone can help ? Please

Accepted Answer

Greg Heath
Greg Heath on 24 Jul 2017
Search both NEWSGROUP and ANSWERS using the search words
Hmin Hmax
The number can be whittled down via
greg Hmin Hmax
Hope this helps.
Thank you for formally accepting my answer
Greg
  2 Comments
lramgi
lramgi on 6 Aug 2017
Edited: lramgi on 6 Aug 2017
Hello,
I am very confused, If you can help.... Where should I see the MSE of the training or the network overall in performance (performance = perform(net,targets,outputs) ) or in mu?
Thank you.
Greg Heath
Greg Heath on 16 Aug 2018
In my posts I use the normalized (in [0,1]) value
NMSE = mse(target-output)/MSEref
where the reference MSE is the value obtained when the answer is assumed to be a constant. Then the minimum MSE is just the average variance of the target.
MSEref = mean(var(target',1))
Hope this helps
Greg

Sign in to comment.

More Answers (0)

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!