Application of PSO-based clustering algorithms on educational databases
Pedro Mendonça dos Santos, Mariana Gomes da Motta Macedo, Elliackin M. N. Figueiredo, Clodomir Santana, Fabiana Soares, Hugo Valadares Siqueira, Alexandre Magno Andrade Maciel, Anuradha Gokhale, Carmelo J. A. Bastos-Filho · 2017
This paper performs an investigation on the application of well-known Particle Swarm Optimization based algorithms for clustering tasks, namely, PSOClustering, Hybrid PSOClustering with K-means and Particle Swarm Clustering. The case study is to tackle a significant problem related to grouping students of an on-line educational database, aiming to increase their learning process through the recommendation of specific grammar lessons. The clustering process is performed based on the type, and the number of errors done by the students. The results show that the PSOClustering algorithm can achieve the best performance when compared to the other PSO-based algorithms and the standard K-means algorithm.