Interleaving Guidance in Evolutionary Multi-Objective Optimization

In this paper, we propose a framework that uses localization formulti-objective optimization to simultaneously guide anevolutionary algorithm in both the decision and objective spaces.The localization is built using a limited number of adaptivespheres (local models) in the decision space. These spheres areusually guided, using some direction information, in the decisionspace towards the areas with non-dominated solutions. We use asecond mechanism to adjust the spheres to specialize on differentparts of the Pareto front by using a guided dominance technique inthe objective space. Through this interleaved guidance in bothspaces, the spheres will be guided towards different parts of thePareto front while also exploring the decision space efficiently.The experimental results showed good performance for thelocal models using this dual guidance, in comparison with their originalversion.