A Student Information System with Students At-Risk Prediction Using Machine Learning for Higher Education Institutions

In today's higher education landscape, identifying and supporting at-risk students is essential for fostering student access and retention. This effects the university's ranking, reputation, and financial well-being. This capstone project addresses the need for an advanced student information system capable of predicting students at risk using machine learning techniques. the main objective is to develop and implement a predictive model within the developed student information system, leveraging machine learning algorithms to identify students who may be at risk of academic failure or dropout and retention. To ensure precise prediction capabilities, the researcher conducted a comprehensive comparative analysis, utilizing four classification models: Artificial Neural Network, Support Vector Machine, K-Nearest Neighbor, and Decision Tree, to gauge the efficacy of identifying at-risk students. The project methodology involves data collection, preprocessing, feature engineering, and model training. The predictive model is then integrated into the student information system using the Flask framework, allowing for real-time identification of at-risk students based on various academic and demographic factors. Bases on the evaluation results performed on the classification models. it became evident that the decision tree model emerged as the most accurate and precise among the selected models. This study contributes to the advancement of student success initiatives of Southers Luzon State University Lucena. The system enables the early detection of at-risk students, serving as a valuable insight into devising and implementing policies to ensure student success.