Comparative Analysis of Classification Algorithms for Student Performance Prediction

Abstract

Student performance prediction is a significant application of data mining in the education sector, aimed at identifying students at risk of poor academic outcomes. Early prediction enables institutions to implement timely interventions and personalized support strategies. This study presents a comparative analysis of four widely used classification algorithms — Decision Tree, Random Forest, K-Nearest Neighbors (KNN), and Naive Bayes — for predicting student academic performance using historical academic and demographic data. The dataset includes attributes such as previous grades, attendance, study time, and socio-economic factors. Standard preprocessing techniques, including data cleaning, normalization, and encoding, were applied before model training. The performance of each classifier was evaluated using Accuracy, Precision, Recall, F1-Score, and Confusion Matrix, along with k-fold cross-validation to ensure reliability. Experimental results indicate that ensemble methods such as Random Forest achieve higher predictive accuracy, while simpler models provide computational efficiency. The findings highlight the importance of algorithm selection in educational data mining and demonstrate the potential of machine learning models to support data-driven academic decision-making.