Dear All,
If simulators for GA optimization for a threshold and later experiments should be different, if so, what difference?
What I know about GA is that fitness function (FF)is only components what define or interact with an environment. it is used to define maxima or minima for best performance. That is used in experiments in the same simulator. That is ideal case, however, this seems to 'copy' pre-determined performance from optimization to experiments.
In order to counter this, do we need different simulators for ga-optimization and experiments? If so, in where or in what terms both should be distinguished.
Hi Muhammed,
Here is link of a paper which is an answer of your question.
-THE USE OF SIMULATIO AND GENETIC ALGORITHM WITH DIFFERENT GENETIC OPERATORS
TO OPTIMIZE MANUFACTURING SYSTEM
www.acs.pollub.pl/pdf/v9n1/3.pdf
Best regards
Dear Mohamed,
Thanks for suggesting this paper. Let me read and come back. It seems this paper uses different simulators based on different genetic operators. It also appears that GA simulators with different GA operators (crossover and selection method) are different simulators. How about GA with different fitness function? are they different simulators as well?
However, What I am interested in different between two simulators one with GA optimization and other used for experiments.
Genetic algorithms have their best performance when playback is done in half the chromosome, keeping constant initial population, little yet known if we make the population decay increasing population mortality, thus there will be better chromosomes that generate solutions faster.
- Badges
- Science topic
- Similar topics
- Computer Science