TurboGA is an augmented version of the Matlab script SpeedyGA (also available on File Exchange). TurboGA uses a simple mechanism called clamping, which, according to a new theory about the workings of genetic algorithms (see http://www.cs.brandeis.edu/~kekib/dissertation.html ), should significantly improve the quality of the solutions returned, especially on large problems.
One of the most effective ways to test this new theory is by studying the efficacy of clamping on a range of problems. That's where you come in. I ask that you report on the efficacy of clamping on the problem(s) that you apply this script to. Please leave your comment on this page, or email it to email@example.com.
Trying this out for the Konka equation y=x1^2+x2^2-3x1-x1x2+3 in the range of (-1..3, -2..2), this optimizer does not finalize at the minimum (2,1), but at (3,2). Hereby (3,2) is the normalized (1,1) and the actual minimum (2,1) would be (0.75,0.75)
Why does the optimizer only try out Booleans 0 or 1, but no real values between 0..1?
Potentially of great help, but what about providing an example of application?
How to optimize like matlab ga.
Please see for example
Changed description. No changes to code.
Updated contact email. No change to code.
Inspired by: SpeedyGA: A Fast Simple Genetic Algorithm
Create scripts with code, output, and formatted text in a single executable document.