AN EXPLORATORY REVIEW OF ARTIFICIAL INTELLIGENCE FOR PREDICTING STUDENTS’ ACADEMIC PERFORMANCE

Abstract

The rapid growth of Artificial Intelligence in education has substantially transformed traditional techniques of monitoring and increasing student academic performance. This study does a comprehensive assessment of research published from 2019 to 2025 about the application of artificial intelligence and machine learning techniques to predict student academic achievement. The main objective was to ascertain the most commonly utilized predictive approaches, examine the most significant student-related attributes, and determine the algorithms that provide the greatest predicted efficacy. A systematic literature review process was conducted using peer-reviewed studies indexed in Scopus, IEEE Xplore, Web of Science, and Google Scholar. To ensure methodological rigor and relevance, research was assessed based on established inclusion and exclusion criteria. The findings demonstrate that supervised machine learning algorithms, such as Naïve Bayes, Artificial Neural Networks, Random Forest, Decision Trees, and Support Vector Machines, dominate contemporary academic predictive research. Ensemble models and neural networks frequently shown enhanced predictive accuracy, especially when trained on extensive and behaviorally diverse datasets. The review shows that a student's academic background and how they engage with learning tools, especially their activity on learning management systems, are the best indicators of their performance, while demographic factors have little effect on how accurate the models are. Substantial obstacles persist, including concerns over data quality, algorithmic bias, restricted interpretability, and the ethical control of student data, despite robust prediction. The research indicates that Artificial Intelligence-based prediction systems have sufficiently advanced to facilitate institutional planning, personalized learning, and early warning systems, contingent upon their implementation within stringent data governance frameworks, transparency, and equity. To enhance the reliability and trustworthiness of educational environments, forthcoming research must focus on the advancement of explainable Artificial Intelligence models, cross-institutional validation, and fairness-conscious modeling.