A Hybrid Particle Swarm Optimizer for Curriculum Sequencing Problem

Curriculum sequencing problem is crucial to e-learning system, which a NP-hard optimization and commonly solved by swarm intelligence. As form of intelligence, particle (PSO) widely used in various kinds problems. However, PSO found ineffective complex The main reason that diversity preservation, leading high risks be trapped the local optima. To solve this problem, novel hybrid algorithm proposed study. First, competitive-genetic crossover strategy for balance convergence diversity. Second, an adaptive polynomial mutation introduced further improve its preservation ability. Furthermore, curriculum scheduling model proposed, where several constraints are taken into considerations ensure practicability sequencing. numerical comparison experiments show effective solving function popular variants; furthermore, designed shows significant advantages over compared algorithms with respect degree satisfaction objectives, i.e., 20, 14, 5 percentages higher, respectively.