Academic performance evaluation using soft computing techniques

This article presents a study of academic performance evaluation using soft computing techniques inspired by the successful application of K-means, fuzzy C-means (FCM), subtractive clustering (SC), hybrid subtractive clustering-fuzzy C-means (SC-FCM) and hybrid subtractive clustering-adaptive neuro fuzzy inference system (SC-ANFIS) methods for solving academic performance evaluation problems. Modelling of students’ academic performance is a difficult optimization problem. We explore the applicability of K-means and FCM, SC, hybrid SC-FCM and SCANFIS clustering methods to the new student’s allocation problem, which allocates new students into some classes that consist of similar students and the number of students in each class not exceeding its maximum capacity. The models were combined with fuzzy logic techniques to analyse the students’ results. In this article, we have conducted clustering based computational experiments to analyse the effects of the different clustering algorithms like K-means, FCM, SC, hybrid SC-FCM and hybrid SC-ANFIS clustering methods for modelling students’ academic performance evaluation. Based on the comparison of the results, it is found that the hybrid SC-ANFIS clustering is better than the other methods.