Leveraging Artificial Intelligence for Enhanced Educational Outcomes: Predictive Modeling and Behavioral Analysis Using the National Education Data Repository

Abstract

This study investigates the application of machine learning to predict academic performance using a synthetic dataset representative of educational outcomes. By analyzing features such as attendance, past academic performance, study hours, and behavioral scores, we developed a regression model capable of accurately forecasting student success. In addition, we explored the model’s classification performance for identifying students at risk of underperformance. Key findings include strong positive correlations between past and current academic performance, highlighting the relevance of historical data in predictive modeling. The model achieved high accuracy in both regression and classification tasks, with an Area Under the Curve (AUC) of 0.89 in classification, indicating robust specificity and sensitivity. These insights underscore the potential of machine learning to enhance data-driven decision-making in education, enabling early intervention and personalized support for students. Further research could explore deeper models and real-world datasets for improved accuracy and generalizability.