Soft Computing Hybrid System for Student Performance Evaluation

Abstract

Education Institutions have deployed technology accelerated learning systems and innovations for effective learning outcomes. Evaluating student’s performance in these systems must align with the cognitive, affective, and psychomotor learning domains. In this research, a Hybrid soft computing system comprising of the Clustering Algorithm, Machine learning technique, and Optimization algorithm were hybridized and implemented to evaluate student academic performance using academic, social, and economic data of students. The quality of Categorizing information first utilizing Fuzzy C-Means and preparing ANFIS utilizing Particle Swarm Optimization was introduced which formed the Hybrid soft computing system (FCM-PSOANFIS). It demonstrated significantly, a robust predictive capability compared to other hybrid machine learning algorithms such as ANFIS and GANFIS. The results of the proposed Hybrid Soft Computing model (FCM-PSOANFIS) show a higher convergence when compare with ANFIS and GANFIS. The proposed model works better with bigger datasets than with smaller or fewer datasets, and it delivers higher predictive findings under settings that depict student learning capacities while assessing student academic achievement.