KESATRIA Reading Tools

Particle Swarm Optimization (PSO) is a population-based optimization which include the use of cognitive and social terms. The cognitive term is represented with the variable of c₁ while social term is represented with the variable of c₂. Both values can be assigned between 0 and 1. The contribution of this research is to compare which role is superior in the Binary Particle Swarm Optimization (BPSO) metaheuristic with Dynamic Increase Cognitive Decrease Social (DICDS) and Dynamic Decrease Cognitive Increase Social (DDCIS) methods, as well as its implementation in the Modified Multi-Objective Agent-Based Hyper-Heuristic (MOABHH). The experiments were carried out 30 times on data set 2 from [1]. The result is that the DDCIS method is 0.4% better in objective value than the DICDS method. This is also proven with the average of number of solutions in the DDCIS method which is more 2.3 solutions than the DICDS method based on the evaluation results carried out by Modified MOABHH. In addition, Modified MOABHH which is run simultaneously with the DICDS and DDCIS methods provides better objective value results of 0.6% compared to the average of both results for each of these methods which are run separately.