Hi Naji,
Yeah, sometimes it is better, but global optimization can be very costly. 'Mountain climbing techniques', are often used because the global optimization problem is very very hard in that case and sometimes because it does not matter. For instance, the techniques used to maximize the likelihood function is a mountain climbing technique, but I thought I remembered that the likelihood is known to have just one maximum if your are estimating a generalized linear model (e.g. logit, probit, normal, poisson, or gamma regression). So any maximum you find will be the global maximum.
Maarten
-----Original Message-----
From: [email protected] [mailto:[email protected]]On Behalf Of Naji
Sent: dinsdag 26 april 2005 13:54
To: [email protected]
Subject: st: Global Optimization
Hi all,
I heard (recently) of Global Optimization.
Advice on lectures (practical oriented : understand it, is it better than
'mountain climbing techniques).
Best regards
Naji Nassar
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
