Bug 247 - Use Simulated Annealing with GA
Summary: Use Simulated Annealing with GA
Status: RESOLVED REMIND
Alias: None
Product: Spamassassin
Classification: Unclassified
Component: Score Generation (show other bugs)
Version: unspecified
Hardware: Other other
: P2 enhancement
Target Milestone: ---
Assignee: SpamAssassin Developer Mailing List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2002-04-23 14:57 UTC by Matthew Cline
Modified: 2002-06-09 16:47 UTC (History)
0 users


Attachment Type Modified Status Actions Submitter/CLA Status
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Matthew Cline 2002-04-23 14:57:54 UTC 
An optimization algorithm which can complement GA is simulated annealing.
When I did a GA project at college, I found that cominging it with annealing
improved the results a lot.

Some overviews of annealing can be found at:

- http://www.cs.sandia.gov/opt/survey/sa.html
- http://www.npac.syr.edu/REU/reu94/ramoldov/proposal/section3_2.html

A parallel annealing library can be found at:

- http://www.uni-paderborn.de/fachbereich/AG/monien/SOFTWARE/PARSA/

A non-parallel library can be found at:

- http://www.taygeta.com/annealing/simanneal.html
Comment 1 Craig Hughes 2002-04-23 15:07:12 UTC 
Subject: Re: [SAdev]  New: Use Simulated Annealing with GA

Matt, if you're interested in the GA code, you should take a look at how it's
working currently.  Rather than using annealing (which admitedly might well
improve the performance -- I haven't tried it yet), it tries to do something
aking to swarming -- take a look at the myMutation function in craig-evolve.cxx

http://www.mkp.com/books_catalog/catalog.asp?ISBN=1-55860-595-9 among others

It's pretty naive the way it's implemented now, but it works really really well,
so why fuck with it is my motto.

C
Comment 2 Craig Hughes 2002-04-23 15:08:16 UTC 
Subject: Re: [SAdev]  New: Use Simulated Annealing with GA

Matt, if you're interested in the GA code, you should take a look at how it's
working currently.  Rather than using annealing (which admitedly might well
improve the performance -- I haven't tried it yet), it tries to do something
aking to swarming -- take a look at the myMutation function in craig-evolve.cxx

http://www.mkp.com/books_catalog/catalog.asp?ISBN=1-55860-595-9 among others

It's pretty naive the way it's implemented now, but it works really really well,
so why fuck with it is my motto.

C


_______________________________________________
Spamassassin-devel mailing list
Spamassassin-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/spamassassin-devel
Comment 3 Craig Hughes 2002-06-10 00:47:18 UTC 
Closing this for cleanup purposes; improvements to the GA are in the pipeline.