[galib] problem with parallel GA in GALib
nguyenvan.lam at ait.ac.th
Wed Feb 9 20:52:05 EST 2005
I have a problem when use parallel GA (I used GADemeGA) with different objectives. I used different objectives for populations in island model.
It means that i assigned each individual objective for one pupulation.
I used GARandomInt for initializing populations to create arrays of integer numbers between lower-bound and upper-bound value.
I didnot derive a new genome because I just use 1DArray build-in genome for my work.
My problem is: after initializing process there also some UNDESIRED CHROMOSOMES WITH ALL GENES' VALUES ARE ZERO that are not in the range of my [lower-bound and upper bound]. Problem about apprearance of strange, undesired chromosmes is disappeared when I tested with other GAs such as: simple ga or steadystate (which only deal with single popolation)
I don't know where strange come from and how to eliminate these violated schromosomes.
Anyone can help me know what is wrong and how to solve the problem?
your help is highly appreciated. Thank you in advance.
Email: nguyenvan.lam at ait.ac.th
forest_jobs at yahoo.com.au
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the galib