<![CDATA[Students are an integral part of higher education institutions, where graduation rates serve as a key indicator of academic quality and institutional effectiveness. To maintain accreditation and academic standards, universities must optimize student graduation rates. Evaluating the factors influencing graduation is crucial in identifying patterns and key determinants that contribute to academic success. This study aims to predict student graduation using Machine Learning, specifically the C5.0 Decision Tree algorithm. The findings indicate a high reliability in predicting student graduation, with an accuracy of 91.35%. The model's ability to identify on-time graduates is reflected in a recall of 93.85% for the On-Time category and 87.18% for the Delayed category. The prediction accuracy is further demonstrated by a precision of 92.42% for the On-Time category and 89.47% for the Delayed category. The F1-Score, which represents the balance between recall and precision, reaches 93.12% for the On-Time category and 88.32% for the Delayed category. These evaluation metrics indicate that the C5.0 algorithm effectively classifies students based on their likelihood of graduating with high accuracy. The predictions generated can serve as a reference for universities to identify at-risk students early, allowing the implementation of appropriate academic strategies to improve graduation rates, accreditation, and institutional quality.]]>
Copyrights © 2025
