The Role of AI In Predicting Students' Academic Grades

Abstract

Students' academic performance is crucial to educators, parents, and researchers. It is a key basis for evaluating education quality and making decisions. It is necessary to understand and predict performance accurately. This study focuses on the prediction of students' academic performance and comprehensively explores the factors affecting performance and prediction methods. The factors influencing academic performance were analyzed from multiple perspectives, including personal perspective, learning motivation and time management, and family and teacher factors. In terms of prediction methods, an overview of machine learning and deep learning prediction models is given, such as Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), Back Propagation Neural Networks (BPNN), etc. These methods all show good prediction results. The study found that multiple factors are intertwined to form a complex performance-influencing mechanism. Although existing prediction methods are diverse and effective, they still face problems such as difficulty in obtaining data sets, high costs, incomplete data, and insufficient model adaptability, which can easily lead to prediction bias. This study aims to improve prediction accuracy, serve educational and teaching practices, and promote the improvement of education quality by optimizing data set utilization, developing adaptive models, and interdisciplinary collaboration.