Welcome, Guest. Please Login
YaBB - Yet another Bulletin Board
 
  HomeHelpSearchLogin  
 
Optimization Methods (Read 5762 times)
Dr. Lulz
YaBB Newbies
*
Offline


Curious HyPhy user

Posts: 9
Optimization Methods
Sep 4th, 2013 at 9:41am
 
Hi, I'm a big fan of HyPhy.

I've been developing models and testing whether how well they fit reality by relatively comparing the log likelihood values.

This said, I'd been using optimization method 4 this whole time and would like to know whether there is a significant difference the methods. It seems as though (given my model) method 0 and 4 give the best values, while I noticed that some optimization methods take significantly less time (methods 5/6).

Is there any literature that I can read on this? I don't seem to see it in the manual (unless I have the truncated version?).

Please let me know!
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Optimization Methods
Reply #1 - Sep 5th, 2013 at 1:56pm
 
Hi there,

The only robust optimization methods are 0 and 4; the rest are experimental and or deprecated and may not even converge, so please don't use them for real data analysis.

Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
Dr. Lulz
YaBB Newbies
*
Offline


Curious HyPhy user

Posts: 9
Re: Optimization Methods
Reply #2 - Sep 6th, 2013 at 9:12am
 
Thank you very much!
Back to top
 
 
IP Logged