Newsgroups: comp.constraints
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!jabowery
From: jabowery@netcom.com (Jim Bowery)
Subject: Re: Feasibility problem with simple non-linear constraints
Message-ID: <jaboweryD4otHM.CAs@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
X-Newsreader: TIN [version 1.2 PL1]
References: <3il0ma$s5h@disunms.epfl.ch> <3ipte2$3gs@gap.cco.caltech.edu> <jaboweryD4nF8v.1K8@netcom.com>
Date: Tue, 28 Feb 1995 01:58:34 GMT
Lines: 19
Sender: jabowery@netcom5.netcom.com

I just realized the probable source of difficulty in my use of ASA.  
About half of my parameters are categorical (integers that represent some 
entirely different set of conditions in the solution without regard to 
numeric ordering).  Therefore, I use the integer, non-derivative mode 
(value of 2 in the typing field of the parameter descriptor lines).  I 
have tried it with the other type settings (-2, -1, and 1 as well as 2) 
but the suboptima(s) keep cropping up as the final answer despite the 
fact that the log shows much better optima's being visited on almost 1/3 
to 1/2 of the runs.

Although I don't know if this matters, I also am using categorical 
parameters that have, at most, 3 categories per parameter (integers can 
take on values from 0 thru 2).

I suspect genetic algorithms would be a better approach in my case.
-- 
The promotion of politics exterminates apolitical genes in the population.
  The promotion of frontiers gives apolitical genes a route to survival.
                 Change the tools and you change the rules.
