Statistical Analysis of Intelligent Systems in Student Performance Evaluation

Abstract

Among various goals of educational institutions, some are to improve the level of education, create as well as execute methods that can benefit students, and evaluate student performance through the school year, at completion of the academic year or in the coming years. Artificial intelligence (AI) methods are employed in this discipline to accomplish desired objectives because of their special capacity to build associations and produce correct outcomes. But the variety of publications and the variation in the information they provide lead to misunderstanding and lessen their capacity to lead new research. Therefore, the conducted research work presents a thorough evaluation of the published papers in research that assess student performance from 2019 to 2026. A quantitative analysis is presented by comparing the considered published scholarly articles based on objective, database, dataset size, method used and accuracy of the developed student performance evaluation model. According to the result of this evaluation, there are more studies being conducted in this field, and a wide variety of AI methods are being used. Moreover, the results show that while issues like data privacy and interpretability still exist, DL and hybrid systems perform better than conventional methods in terms of predicted accuracy. However, the research that is now available indicates that AI can be helpful in detecting and enhancing a variety of educational achievement domains.