Back to Recipes Homepage
recipe picture
Recipe: CG303 GAB V4.70+EO
Your rating: None Average: 4.2 (22 votes)


Name: CG303 GAB V4.70+EO
ID: 42156
Created on: Tue, 06/05/2012 - 17:59
Updated on: Wed, 06/06/2012 - 00:59

Feature request by jeff101. Now, we can not only do Genetic Algorithm but also Extremal Optimization. This will skip breeding and just reuse all good band clusters (also those which would be overwriten by breeding) but still replace bad clusters new random ones, as we did in GA, too. To try this method, see information online.

Best For


Joined: 09/18/2009
Groups: SETI.Germany
Code as text... here:

Do use Extremal Otimization instead of genetic Algorithm, change code in line 1515, which is:




Maybe you want to compare both strategies.
Then save the puzzle before, and look at line 1654:


change it to

RNDseed=[here a constant value]
so that the random numbers used will be the same in both runs.

Then you can run the script at first in one mode, reload your puzzle state, change the mode and run it a second time.

jeff101's picture
User offline. Last seen 1 hour 9 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
References about Extremal Optimization discusses Extremal Optimization. is a 2009 Master's Thesis that uses it.
pp.10-11 of this Thesis discuss comparisons between Extremal Optimization and the Genetic Algorithm. has a description of Extremal Optimization under Research/Optimization.

jeff101's picture
User offline. Last seen 1 hour 9 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
One of the above references has moved:

The 2009 Master's Thesis by Eric Drucker above has moved to:

Want to try?
Add to Cookbook!
To download recipes to your cookbook, you need to have the game client running.



Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons