Comparative Analysis of Sequencing Methods and Markov Models for Predicting High-Achieving Students at Budi Darma University

Abstract

The prediction of high-achieving students is a strategic step in supporting the development of academic quality within higher education institutions. This study aims to compare two data mining approaches, namely the Sequencing method and the Markov Model, in predicting high-achieving students at Universitas Budi Darma Medan. The Sequencing method is used to identify patterns in the sequence of academic grades and non-academic activities of students from semester to semester, while the Markov Model is used to calculate the probability of transitions in students' academic status based on historical data. The research adopts a quantitative approach involving 100 active students with complete academic and non-academic data. The data analyzed include semester GPA, participation in organizations, seminars, and achievements in competitions. Both methods were evaluated using metrics such as accuracy, precision, recall, and F1-score. The evaluation results show that the Sequencing method achieved an accuracy of 87%, precision of 85%, recall of 88%, and an F1-score of 86%, while the Markov Model recorded an accuracy of 81%, precision of 79%, recall of 83%, and an F1-score of 81%. Based on these results, the Sequencing method is considered superior in detecting patterns and providing more accurate predictions of students’ achievement potential. The comparison of these two methods provides a foundation for institutions to develop more accurate, objective, and comprehensive student achievement prediction systems. Thus, universities can implement early and well-targeted interventions and guidance.