A Comparative Analysis of Data Mining Models for Student Performance Prediction in School Education

Authors: Zahira Noor Quraishi, Atul Dattarya Newase

Full-text Research PDF File:   View   |   Download

Abstract: Student performance prediction has become a central focus of Educational Data Mining (EDM) due to its potential to support early intervention and improve learning outcomes. With the increasing availability of educational data, various data mining and machine learning models have been applied to analyze academic, behavioral, and demographic factors influencing student achievement. This study presents a comparative analysis of commonly used data mining models for student performance prediction in school education, with particular emphasis on primary and secondary levels. Drawing upon recent empirical studies and systematic reviews, the research examines traditional machine learning algorithms such as Decision Trees, Random Forest, Naïve Bayes, Regression models, and Neural Networks, alongside pre-processing techniques including feature selection and class imbalance handling. The analysis highlights that ensemble-based models, particularly Random Forest, consistently demonstrate strong predictive performance across diverse datasets, while simpler models offer better interpretability. However, most existing studies rely on higher education or single-institution datasets, limiting their applicability to school-level contexts. By synthesizing methodological trends, performance outcomes, and limitations, this study identifies critical gaps and emphasizes the need for general, explainable, and context-aware models for school education. The findings aim to guide researchers and practitioners in selecting suitable data mining models for effective student performance prediction.

Keywords: Educational Data Mining, Student Performance Prediction, Machine Learning, Primary Education, Secondary Education

Paper Id: 5.513

Published On: 2024-07-07

Published In: Special Issue - International Conference on Trends and Innovations in Management, Engineering, Sciences and Humanities (July 2024)

Share this