Abstract
Education stands as a critical driver of human capital development and sustainable economic progress worldwide, reflected in its inclusion in Sustainable Development Goals. The quality of education, gauged through student academic performance, signifies the effectiveness of formal learning systems. This study compares traditional and modern analytical methods, specifically AI and machine learning approaches, in analyzing student performance in an Economics-based course, Econometrics, at Prince Abubakar Audu University, Nigeria. Using a sample size of 1097 students' demographic and academic data, the study explores 13 features, focusing on Econometrics. Logit regression, representing the traditional method, identified five significant predictor variables with limited accuracy. Conversely, machine learning algorithms, including Artificial Neuron Network (ANN), Naïve Bayes, K-Nearest Neighbor (KNN), Random Forest, and Support Vector Machine (SVM), outperformed logit regression. SVM, in particular, exhibited the highest accuracy (85.48%) and precision. Given the evolving complexity of educational processes, the study advocates for the widespread adoption of AI-enhanced machine learning techniques in education data analysis to ensure reliable insights and informed decision-making for qualitative learning systems in Nigerian higher institutions.
Keywords: Education Data Mining, Traditional Analytical Methods, Machining Learning Classifiers, Tertiary Institution, Nigeria