Machine learning based Innovative System to Forecast Secondary School Students' Academic Accomplishments

Abstract

Predicting secondary school students' academic accomplishments is crucial for early intervention and personalized learning strategies. This study develops a machine learning-based system to forecast student performance, including grade and percentage prediction, while analyzing the impact of various socio-economic, educational, personal, and technological factors. The dataset was collected through a structured survey, incorporating aspects such as family income, parental education, access to private tutoring, school infrastructure, learning environment, mental health, career guidance, geographic constraints, government policies, and digital literacy. Pre-processing was done using five machine learning algorithms: Random Forest, Support Vector Machine, Decision Tree, K-Nearest Neighbors, and Gradient BoostingTo assess model performance, various evaluation metrics, such as accuracy and root mean squared error, were utilized. The findings suggest that machine learning methods are capable of accurately forecasting student performance, with the Random Forest algorithm demonstrating the greatest level of precision. This study lays the groundwork for AI-based educational resources aimed at recognizing students who are at risk and facilitating focused interventions.