Loading...

Help With Adaptive Simulated Annealing (ASA) Optimization

977 views

Loading...

Loading...

Loading...

Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on May 9, 2014

This Adaptive Simulated Annealing (ASA) video outlines the motivation behind ASA.

Many systems require fitting models to data, or require sampling the space of alternative variables/parameters. A class of algorithms that is robust for many systems is importance-sampling. A subset of this class is simulated annealing (SA). SA algorithms -- only if applied correctly -- come with a mathematical proof that optimal solutions can be found with statistical certainty. ASA, originally called Very Fast Reannealing (VFSR), has been available and maintained at no charge to anyone at several websites since 1989.

The code can be downloaded from
http://www.ingber.com/#ASA
http://alumni.caltech.edu/~ingber/#ASA

The best guide for ASA is in the ASA-README.[] file that comes with the code, which also is available separately at the above websites in pdf, txt, html, and gzip'd ps formats. Especially read Sections
"Use of Documentation for Tuning"
"Simulated Annealing is NOT Simulated Quenching"
There are several published papers also available from the above sites.

The code has a steep learning curve, and you may benefit from finding some local expert to help you. If you get stuck, I can consult/contract with you to get started or even work with you towards completion of your optimization/importance-sampling project. See
http://www.ingber.com/ingber_terms.html
for more information on this choice.

Note that I regularly address short technical queries via email, at no charge.

Lester Ingber

  • Category

  • License

    • Standard YouTube License
Comments are disabled for this video.
When autoplay is enabled, a suggested video will automatically play next.

Up next


to add this to Watch Later

Add to

Loading playlists...