Comparison of Machine Learning Algorithms for Predicting Student Academic Performance and Stress Level

Abstract: The traditional education system faces significant challenges in analyzing and predicting student performance, especially in regions with large student populations. Institutions often rely on outdated evaluation methods and lack a systematic approach for continuously monitoring academic performance and psychological stress levels. To address these challenges, in this data integrates academic, demographic, psychological, and technology-related factors, providing a more comprehensive view of student behavior and learning outcomes. The model evaluates and compares Machine Learning (ML) algorithms likes Naïve Bayes, Decision Tree, K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Random Forest to determine the most effective technique for predicting student performance and stress levels. As a result, identifying academically weak and mentally stressed students becomes difficult, leading to delayed interventions. The Performance analysis is conducted for ML algorithms based on classification accuracy and compared with existing research work, demonstrating improved prediction capability and offering valuable insights for educational decision making. Keywords: Machine Learning, Educational Data Mining, Random Forest, Support Vector Machine, K-Nearest Neighbors, Naïve Bayes, Decision Tree