Experimental analysis of dynamic migration intervals on 0/1 knapsack problems

In this paper, the effects of adapting the migration intervals on the performance and solution quality of dynamic multi-population genetic algorithms (MGAs) are investigated. An adaptive scheme to evolve the appropriate migration intervals for MGAs is presented. An adaptive multi-population genetic algorithm which can adjust the migration intervals is proposed. Experiments on the 0/1 knapsack problem are conducted to show the effectiveness of the proposed approach. The experimental results have illustrated the effectiveness of the self-adaptation of migration intervals for MGAs.