Using machine learning to analyze factors influencing grades in upper-level information technology courses

Understanding the determinants of academic success in upper-level Information Technology (IT) courses is critical for improving student outcomes and informing pedagogical strategies.This study explores the predictive relationships among attendance patterns, exam performance, homework scores, and final grades using machine learning techniques.Drawing on data from 162 students enrolled in four advanced IT courses, we employed three classification models-Decision Tree (DT), Artificial Neural Network (ANN), and Nave Bayes (NB)-to assess their effectiveness in forecasting student performance.Results revealed that ANN outperformed the other models with a prediction accuracy of 79.01%, indicating that complex, non-linear interactions among academic factors are influential.DT and NB also surpassed the baseline model, highlighting the relevance of key predictors, especially exam scores and first-day attendance.Our findings underscore a strong link between early attendance and academic achievement, suggesting that absence on the first day of class is a significant early-warning indicator of poor performance.The study concludes with practical implications for early intervention strategies and proposes future research directions involving expanded datasets and cross-disciplinary analysis.