Page 1 of 1
constant gain learning algorithm URGENT!!!
Posted:
Sun Mar 03, 2013 10:01 pm
by alexanderm
I want to implement a constant gain learning algorithm for inflation expectations,but i couldn't find any example on the website on how to implement recursive algorithms in Dynare.
Is it necessary to include the way parameters are update recursively in the declaration of the model?
Could someone help me with an example,please?
Thanks in advance!
Re: constant gain learning algorithm URGENT!!!
Posted:
Sun Mar 10, 2013 10:18 am
by jpfeifer
Sorry, but in this full generality, I doubt that anyone is able to help you. This looks like a really complicated topic.
Re: constant gain learning algorithm URGENT!!!
Posted:
Sat Mar 30, 2013 8:50 pm
by alexanderm
I would like ro represent inflation expectation using an adaptive learning algorithm based on constant gain algorithm.
I don't understand if i have to put in the model the recursive way in which parameters are set ?
Dynare implements Kalman filter?How can I use this procedure?I haven't found an example for this in Dynare tutorial.
Thanks in advance!
Re: constant gain learning algorithm URGENT!!!
Posted:
Fri Apr 05, 2013 6:32 pm
by jpfeifer
This seems to be a very special application. I am pretty sure that Dynare cannot solve it out of the box. If you as the model author are not sure if Dynare can do it, try something else.