A dynamic particle swarm optimizer for educational data processing

Many problems in educational data processing on Internet are dynamic optimization problems. These DOPs challenge traditional optimization methods as well as conventional evolutionary algorithms. In this paper, an improved Particle Swarm Optimizer (PSO) by introducing Group Search Optimizer (GSO) mechanisms is proposed to solve optimization in dynamic environment. The member switch method between two roles in GSO is employed as immigrant method among multi-swarms. Experimental results on unimodal and multimodal benchmarks demonstrate the proposed algorithm can trace the moving optimal and are effective in dynamic environment.