Abstract: Predictive models for student academic success are often hindered by class imbalance, particularly in identifying students at risk of dropping out. This study evaluates nine machine learning ...