termination criterion with NLPQN

madix — Tue, 28 Nov 2017 13:33:49 GMT

Hello,

I am doing a quasi newton optimization method but I have a problem. The optimization stop before I find the global minimum.

I wonder if I could use a termination criterion of 1e-4 in order to stop the optimization after getting an objective function as small as I want and if it is possible, how to do it?

Here is my code

proc iml;

use a;

read all var _NUM_ into data;

close a;

* Chargement des fonctions ;

reset storage = &lib..Functions;

load module=(function1 function2);

parameters = j(5, 1, .) ;

parameters[1] = 1.1 ;

parameters[2] = 5;

parameters[3] = 2;

parameters[4] = 1.5;

parameters[5]= 0.9;

con = {-10 1E-6 -10 1E-6 1E-6, 10 10 10 10 1}

opt = {1 4} ;

*Quasi-Newton Method ;

call NLPQN(rc, parameters, "function1", theta, opt, con);

quit ;

Moreover, this is the value of the objective function I get. You can see it is quite neer from each other at 10-4

Thank you for helping

Re: termination criterion with NLPQN

Ksharp — Tue, 28 Nov 2017 13:18:45 GMT

Post it at IML forum,since it is real IML code. and could try another initial value .

topic Re: termination criterion with NLPQN in SAS Programming

termination criterion with NLPQN

Re: termination criterion with NLPQN