Machine Learning Based Academic Performance Prediction System Considering Socio-Economic and Environmental Influences

Dr. A. S. C. Tejaswini Kone,
M. Sreeja, S. Sai Varsha Vardhini, S. Simhadri, V. Lavaraju, G. L. S. Sowjanya
Department of Computer Science and Engineering
Visakha Institute of Engineering and Technology, Visakhapatnam, Andhra Pradesh, India

Abstract:The rapid advancement of technology and the growing availability of educational data have created new opportunities to apply machine learning techniques in predicting student academic performance. Traditional evaluation methods mainly rely on examination scores and attendance records, which often fail to consider the multiple factors influencing student success. This research presents a Machine Learning Based Academic Prediction System that analyses academic, socio-economic, and environmental factors to generate accurate predictions about student performance and identify students who may be at academic risk. The proposed system collects student data such as previous semester grades, attendance percentage, study hours, assignment performance, family income, internet access, and environmental conditions. These inputs are processed using machine learning algorithms such as Linear Regression, Decision Tree, and Random Forest to predict the next semester Semester Grade Point Average (SGPA) and determine whether a student is at risk or not at risk academically. The system is implemented as a web-based application using Python and Streamlit, providing a user-friendly interface for students and administrators to enter data, view predictions, and generate performance reports.The results demonstrate that the system improves prediction accuracy and supports data-driven decision-making in educational institutions. By enabling early identification of students who require academic support, the system enhances academic planning, reduces failure rates, and improves overall student performance.Keywords: Machine Learning, Academic Prediction, Student Performance, Educational Data Mining, PredictiveAnalytics, Risk Detection