Enhancing Higher Educational Student Performance Using Data Mining Techniques

This study is to explore how students' performance in higher education might be improved by employing educational data mining approaches, with a particular emphasis on students. Data collected over a five-year period (2015–2023) from an educational institution was employed to analyze and predict student performance using three key techniques: ID3, C4.5, and Random Forest. The results indicated that C4.5 achieved higher accuracy compared to the other techniques. The decision tree constructed by C4.5 was efficiently pruned, making it more effective. Furthermore, the performance of the J48 model, which is an implementation of C4.5 in Weka, outperformed both ID3 and Random Forest. The study reveals that employing the C4.5 algorithm for data mining consistently yields superior results in predicting and improving undergraduate student performance compared to alternative methods. It underscores the efficacy of utilizing data mining techniques to enhance educational outcomes in higher education. It highlights how C4.5 algorithm can achieve faster and more accurate results, emphasizing the potential for enhancing educational outcomes.