Advancing Sustainable Educational Practices Through AI-Driven Prediction of Academic Outcomes
Saleh Albahli
Abstract
The integration of artificial intelligence (AI) into educational systems has the potential to transform academic practices and promote sustainability in education. This study explores the development and evaluation of machine learning (ML) models to predict student performance, integrating socio-demographic, academic, and behavioral data to enhance accuracy and interpretability. By leveraging advanced techniques such as convolutional neural networks (CNNs) and explainable AI (XAI), this research provides actionable insights into key factors influencing student success, such as attendance and socio-economic status. The results demonstrate that CNNs achieve exceptional predictive accuracy (99.97%) compared to traditional models, while XAI methods ensure model transparency for informed decision-making. These findings enable the design of personalized learning strategies, timely interventions, and equitable educational practices that contribute to student retention and overall institutional efficiency. This study aligns with the goals of sustainable education by emphasizing data-driven approaches to enhance learning outcomes, equity, and resource utilization.