Set-Based Particle Swarm Optimization for the Multi-objective Multi-dimensional Knapsack Problem

The set-based particle swarm optimization algorithm (SBPSO) was developed to solve the single-objective multi-dimensional knapsack problem. Since then, the SBPSO was applied to a number of, mostly single-objective, optimization problems. Recently, the SBPSO was adapted to solve multi-objective optimization problems (MOPs). The resulting multi-guide SBPSO (MGSBPSO) showed excellent performance on multi-objective portfolio optimization problems and multi-objective formulations of rule induction. This paper applies the MGSBPSO to the multi-objective multi-dimensional knapsack problem (MOMKP) in comparison with state-of-the-art algorithms. The results show that the MGSBPSO performs very competitively.