<![CDATA[Education plays an important role in improving the quality of human resources and supporting a country’s progress toward becoming a developed nation. Higher education institutions serve as one of the providers of formal education, where the quality of these institutions is measured through accreditation. One of the key indicators influencing accreditation is the outcomes and achievements of the Tri Dharma of higher education, which include the timeliness of student graduation. This study aims to compare models for predicting on-time student graduation using three machine learning algorithms, namely Decision Tree, Naïve Bayes, and Support Vector Machine (SVM), as well as their combination through the Ensemble Voting method. The prediction is based on historical grade data from courses taken during semesters one to four. The research methodology adopts the Cross-Industry Standard Process for Data Mining (CRISP-DM), which consists of six stages: business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The dataset used in this study consists of 2,471 records with 11 attributes. Data preprocessing was conducted through data cleaning and class balancing using under sampling techniques. The results indicate that the Ensemble Voting model using the Soft Voting method achieves the best performance, with an accuracy of 91.80%, precision of 91.87%, and recall of 91.80%, outperforming the individual models of Decision Tree, Naïve Bayes, and SVM. The implementation of this model can be utilized to predict students’ on-time graduation based on course grade inputs. Therefore, this research can serve as a supporting tool for early detection of potential delays in student graduation.]]>
