Forming automatic groups of learners using particle swarm optimization for applications of differentiated instruction

Abstract

Abstract The aim of this paper is to present a method that uses computational intelligence techniques to classify students according to the principles of differentiated instruction. A clustering algorithm based on particle swarm optimization is applied to two sets of data emerging from the holistic assessment of the student's particular characteristics and needs. The results illustrate the algorithm's contribution to the effective formation of heterogeneous student groups, with the members of each having homogeneous characteristics of skills, difficulties, psychosocial and cognitive profiles. Thus, the teacher can easily manage students, by knowing the characteristics of each group. A comparison with a genetic algorithm as well as cuckoo search algorithm shows that the proposed method provides improved categorization capabilities.