A Comparative Study of Data Mining Techniques for Student Performance Analysis in Education

Authors: Zahira Noor Quraishi, Atul Dattarya Newase

Full-text Research PDF File:   View   |   Download

Abstract: The rapid growth of digital technologies in education has led to the generation of large volumes of data related to student learning, assessment, and behavior. Educational Data Mining (EDM) has emerged as a powerful approach for analyzing such data to improve academic decision-making and student outcomes. This study presents a comparative analysis of data mining techniques used for student performance analysis across different educational contexts. Based on a synthesis of recent empirical studies and systematic literature reviews, the research examines commonly applied methods including decision trees, random forest, Naïve Bayes, neural networks, clustering techniques, feature selection approaches, and process frameworks such as CRISP-DM. The comparison highlights variations in model performance, interpretability, scalability, and applicability across higher education, secondary education, and mixed learning environments. Findings indicate that ensemble models, particularly random forest, consistently achieve strong predictive accuracy, while simpler models offer better transparency for educational stakeholders. However, the analysis also reveals limitations such as dataset dependency, lack of generalizability, data quality issues, and underrepresentation of school-level education in empirical research. By identifying strengths, weaknesses, and methodological gaps, this study provides guidance for selecting appropriate data mining techniques and supports the development of more robust, explainable, and context-aware student performance analysis systems.

Keywords: Educational Data Mining, Student Performance Analysis, Machine Learning, Data Mining Techniques, Learning Analytics

Paper Id: 7.710

Published On: 2025-01-15

Published In: Special Issue - International Conference on Engineering, Economics, Management and Applied Sciences (January 2025)

Share this