TITLE:
Predicting Student Success through Learning Analytics: A Comparative Modeling Approach
AUTHORS:
Artemis Rigou, Foteini Kyriazi, Dimitrios Thomakos
KEYWORDS:
Forecasting, Learning Analytics, Prediction, Higher Education
JOURNAL NAME:
Creative Education,
Vol.16 No.8,
August
15,
2025
ABSTRACT: In this paper, we explore the application of predictive modeling within the field of Learning Analytics (LA) to forecast student academic success in higher education. Utilizing the Open University Learning Analytics Dataset (OULAD), we integrate student demographic, educational, and assessment data to build a dataset suitable for supervised learning. Two models are employed: logistic regression, chosen for its interpretability, and Random Forest, selected for its capacity to capture complex, non-linear relationships. Our target variable is whether a student passes a course module. The analysis reveals that performance in early assessments is the most influential predictor, followed by prior education level and age group. The Random Forest model consistently outperforms logistic regression across all performance metrics, including accuracy, precision, and recall. These results emphasize the potential of machine learning to support early identification of at-risk students, guiding timely interventions. We conclude by discussing the policy implications of our findings for institutional strategies aimed at improving student retention and academic outcomes.