[Scilab-users] Genetic Algorithm question

[ebmb.sci]

 Hello All,
 I've been used the Genetic Algorithm in Scilab 5.4.1 (with the optim_ga
macro). After a lot of tests, I would like to report a strange behaviour.
That is, from the second time that I ran the algorithm, I always note a time
increment to get the results. For example, to optimize a schedule problem
with 40 tasks, at the first time the results could be given in 5 minutes,
but from the second time it will be much more than 5 min. (This time is only
a example, ok?!)

 I also noted that when I close the Scilab environment and start it again, I
can get my results, for example, in the same 5 minutes. (But, I always will
need close and start to get time and results (in terms of quality)
corrects.)

 So, my questions are:

1. What is happening with my Scilab?
2. Is that behaviour correct?
3. Somebody who works with optimization already had problems like this?

 Thanks in advance for attention!

 Best regards,
 Eduardo.



--
View this message in context: http://mailinglists.scilab.org/Genetic-Algorithm-question-tp4028501.html
Sent from the Scilab users - Mailing Lists Archives mailing list archive at Nabble.com.