In the evolving landscape of higher education, data-driven approaches have become pivotal in enhancing academic performance and institutional decision-making. This study investigates the application of supervised machine learning algorithms to predict undergraduate students’ outcomes in Introduction to Computing Science at the Federal University of Allied Health Sciences, Enugu, Nigeria. The aim is to develop predictive models capable of early identification of students at risk of academic failure, enabling proactive intervention strategies. A dataset comprising 500 anonymised student records, including demographic, behavioural, and academic features, was preprocessed using normalisation and encoding techniques. Feature selection methods, such as Chi-square tests and Recursive Feature Elimination (RFE), identified midterm test scores, attendance rate, and parental education as key predictors. Five classification algorithms, Logistic Regression, Decision Tree, Random Forest, Support Vector Machine (SVM), and Gradient Boosting, were trained and evaluated using 5-fold cross-validation. Results revealed that ensemble models outperformed traditional classifiers, with Gradient Boosting achieving the highest performance (87% accuracy, 0.85 F1-score, and 0.91 ROC-AUC). Feature importance analysis confirmed that early assessments and engagement metrics are strong indicators of final course performance. These findings underscore the potential of machine learning to enhance academic support systems by providing actionable insights for educators and administrators. The study concludes by recommending the integration of predictive analytics into institutional frameworks, the development of academic early warning systems, and future expansion of the model to include behavioural and real-time learning data. This work contributes to the growing field of Educational Data Mining and presents a scalable model for fostering academic excellence in Nigerian higher education.