A Framework of an Intelligent Recommendation System for Particle Swarm Optimization Based on Meta-learning

Particle swarm optimization has shown great advantages to solve NP-hard problems due to its simplicity, intelligence, efficiency and easy enhancement. However, with a large number of particle swarm optimization variants (PSOs) proposed, there are two issues: First, are the general problems of PSOs in terms of premature convergence, universality and robustness solved thoroughly? Second, how to find the relatively appropriate PSOs in a quick and efficient way when facing real-world complex optimization problems? Therefore, it is so necessary to develop an intelligent recommendation system for PSOs to provide users a black-box tool for various application problems.