Parameters suitability for changing the search behavior in adaptive GA

The effect of the parameter values in GA is undeniable. Different working operators as well as different values for the parameters would result in different efficiency and performance for a given problem. The effect of parameters on exploration and exploitation is the main effective factor in GAs. Keeping a balance between the exploration and exploitation during the run would lead to good results. Different parameter has different role in keeping this balance. An adaptive parameter control is proposed. A weight is given to each parameter depend on its power in exploration. The parameters then will be ordered based on their explorative power. A comparison has been made among different orders of the parameters. The results suggest the most suitable parameters in exploration and exploitation in the process of evolution. The results have shown the selection pressure as the most explorative parameter and then mutation and crossover rates has the most effect on exploration.