Beyond Grades: A Machine Learning Approach to Analyze the Impact of Behavioral Patterns on Student Performance

Beyond Grades: A Machine Learning Approach to Analyze the Impact of Behavioral Patterns on Student Performance.

Prof. Priyanka V H1, Harshan A 2, Likith D 3, Raja ul haq4, Sharanyae S S5

1Dept. of Computer Science and Engineering & JSS Science and Technology University

2Dept. of Computer Science and Engineering & JSS Science and Technology University

3Dept. of Computer Science and Engineering & JSS Science and Technology University

4Dept. of Computer Science and Engineering & JSS Science and Technology University

5Dept. of Computer Science and Engineering & JSS Science and Technology University

Abstract - Academic success is not only influenced by tradi- tional factors like classroom performance and attendance, but also by lifestyle and behavioral habits such as sleep duration, screen time, and app usage. This research presents Beyond Grades, a system that leverages machine learning models to analyze and predict student performance using both academic and non-academic variables. Data collected from Kaggle was processed and used to train Decision Tree and Random Forest models. The insights were presented through an interactive Flask and Power BI dashboard, enabling students and faculty to visualize the behavioral factors that affect performance. The system achieved an accuracy of 87.2, offering a practical and scalable approach to early academic support. This study aims to highlight the role of behavioral data in improving educational outcomes.

Key Words:  Machine Learning, Sleep Efficiency, Student Performance Prediction, Flask API, Power BI, Random Forest.