Community Profile

Jorge Amaral


11 total contributions since 2010

Contact

Contributions in
View by

Responded


Re: How to train different narxnets with the same training, validation and test sets
Greg, In my previous question I calculate MSE00s=mean(var(Tstrain',1)) which is equal to MSEtrn00. I understood that for t...

2 years ago | 1395 views

Responded


Re: How to train different narxnets with the same training, validation and test sets
Apologies the message was sent before I finished. I use the code and calculate the R2 for several datasets (open loop) and comp...

2 years ago | 1395 views

Posted


How to train different narxnets with the same training, validation and test sets
Hi, 1) I want to train different narxnets with the same training, validation and test sets. 2) I want to know how to calcu...

2 years ago | 1395 views

Posted


Using crosscorrelation to obtain the delays in a arx process for a narxnet design
Hi, I am using a narxnet to predict the output provided by an arx process in order to understand the use of ARX. The ARX proc...

2 years ago | 1084 views

Posted


When the output is sampled in the Simulation of ARX Process
Hi all, I created an ARX process using the command: process_arx= idpoly([1 -0.5],[0 0 0.6 -0.2],1,1,1,'NoiseVariance',0.05,'...

2 years ago | 953 views

Posted


Is there a simple way to print only nonzero elements of matrix?
Hi, Is there another way to print the nonzero elements in a matrix other that use for loops? For example: a= [ 1 0 3; 0 0 4;...

6 years ago | 3249 views

Posted


Neural network trained with bayesian regularization
Hi, 1) I look at the matlab's example. p = [-1:.05:1]; t = sin(2*pi*p)+0.1*randn(size(p)); net=newff(minmax(p),[20,1],{'tan...

6 years ago | 2769 views

Responded


Re: Using correlation to select inputs
SNIP Hi Greg, I forgot to put in the other post that the STEPWISEFIT with backward seach did generate a better feature vector...

7 years ago | 2682 views

Responded


Re: Using correlation to select inputs
SNIP rank(Xa) ans = ans = Looking at your post about pre-training advice, cond(Xa) should around 100, so it looks like ...

7 years ago | 2682 views

Responded


Re: Using correlation to select inputs
SNIP Greg, Thanks, for you answer. I followed part of your suggestions below: x=zscore(x); y is already 0 or 1; Xa = [on...

7 years ago | 2682 views

Posted


Using correlation to select inputs
Hi, When I use a one nearest neighbor classifier in 5 fold crossvalidation. I got better results using just using input 9 and...

7 years ago | 2682 views