Machine Learning-Based Student Performance Prediction Using Socio-Demographic and Academic Features

In recent years, educational institutions have increasingly turned to data-driven decision-making to enhance student outcomes. This research paper presents a machine learning approach for predicting student academic performance using socio-demographic and academic features. The study leverages publicly available data from secondary school students, incorporating attributes such as study time, past grades, parental background, and lifestyle factors. Several classification algorithms—including Logistic Regression, Decision Tree, Random Forest, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN)—were applied and compared based on accuracy and precision metrics. The results highlight the Random Forest classifier as the most effective model, achieving the highest accuracy in predicting final grades. The findings can support educators and policymakers in identifying at-risk students and providing timely interventions. This study demonstrates the potential of machine learning to foster academic success through early prediction and personalized support.