According to this site (http://www.swarmintelligence.org/tutorials.php) c1 and c2 in the main equation of PSO are learning factors and c1=c2=2.
Does this hold true for solution? I hope not. Then what are the methods to learn it optimally or nearly-optimally?
U should follow my research papers. Which help you to solve ur problem.
I appreciate your involvement on this Sharandeep but I would really be grateful if you could take the time and specifically point towards my question.
Thanks in advance.
You can try different values of the coefficients. Varying the magnitude of the coefficients, you have different behaviour of the swarm. According to my experience, there is not a general rule that applies to all the optimization problems. Each problem has its own characteristics and its own convergence properties. As a consequence, the coefficients can be 'optimized' in order to help convergence. However, to the best of my experience, finding the optimal coefficient is just a trial & error procedure.
it is random value . according to issue set value .
in the one issue set value 1.5
and another issue value is 2.2 .
so you must train and test difference value to rich aim.
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning