Predicting student risk in higher education using machine learning and institutional data analytics: a Southeastern European perspective

Key Points

• The aim is to explore how machine learning can predict student risk in higher education.
• Utilized institutional data for analysis
• Employed machine learning algorithms
• Assessed student outcomes based on predictive results
• Achieved a prediction accuracy of over 85% for identifying at-risk students
• Identified key factors such as attendance and grades as predictive metrics
• Provided actionable insights for intervention strategies