Machine Learning Approaches to Predicting Student Academic Failure in Nigerian Distance Learning Institutions

Student attrition and academic failure in Nigerian distance learning institutions represent significant resource wastage and human capital losses, yet early identification of at-risk students through data-driven methods remains underdeveloped in the sector. This study evaluates machine learning approaches for predicting student academic failure in Nigerian distance learning institutions, using learner data from the National Open University of Nigeria. A dataset comprising anonymised academic records of 4,800 enrolled students over four academic sessions was used, with features including engagement metrics, assignment submission timeliness, prior academic performance, and demographic variables. Five machine learning algorithms were evaluated: logistic regression, decision tree, random forest, support vector machine, and gradient boosting. Model performance was assessed using accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve. The gradient boosting model achieved the highest predictive performance, with an F1 score of 0.84 and an area under the curve of 0.91. Assignment submission regularity and mid-course assessment scores were identified as the most predictive features. The study addresses a significant research gap in the application of learning analytics to the Nigerian distance education context and proposes an early warning system architecture based on the best-performing model. Ethical considerations regarding student profiling are discussed. Keywords: machine learning, student failure prediction, distance learning, National Open University of Nigeria, learning analytics.