Predictive Forecasting of Students’ Performance Utilizing Machine Learning Techniques

Predicting student performance is essential for early intervention and tailored educational support, enabling educators to address challenges proactively and enhance student outcomes. The goal is to predict the student's academic performance based on their individual information and a variety of academic and non-academic criteria. Prior studies have predominantly focused on academic indicators, with limited exploration of non-academic variables. Previous research utilizing machine learning techniques and data mining techniques has shown promise but lacks comprehensive integration of diverse influencing factors. Employing the machine learning algorithms, it aims to uncover complex patterns within the data. The inclusion of socio-economic factors and extracurricular activities aims to provide a holistic perspective on student performance. Logistic Regression achieved an accuracy of 90.76% with precision rates exceeding 91% across performance categories. Naive Bayes attained an accuracy of 89.08%, with precision values ranging from 83% to 93%. Decision Trees exhibited respectable precision values ranging from 85% to 88%. The anticipated outcome is the creation of a web application that integrates the best-performing predictive model, offering educators and administrators a tool for personalized student support. This application aims to enhance academic performance by providing insights in educational settings.