Prev: DCC GARCH
Next: using fftn and ifftn to compute distribution of multivariate Student's t
From: Nicola on 18 Sep 2009 10:22
Hi guys!
I have done an experiment of identification using command line in matlab. I have good result from the crossvalidation on the validation date. The fitting is more or less 85%.
My problem is that when i use a idpoly object in simulink via idmodel block, the output of the sistem on the same set of date of the validation process is very diffent and no good!
can somebody help me?

I really dont understand why that huppened!

NB. If i do the same experiment via nonlinear identificazion (nlarx) and then i use idnlarx model block of simulink i have not problem!
 | 
Pages: 1
Prev: DCC GARCH
Next: using fftn and ifftn to compute distribution of multivariate Student's t