In this article, we developed a model for predicting graduation rate and college students performance. Predicting students performance is a necessity for a university or faculty. If from the beginning, the faculty can estimate how good or bad their students in learning process, then they can give the right treatment so they can graduate on time with good GPA. Data used in this model was alumni data from Universitas Pekalongan in year of 2018 which consist of 1130 alumni. The model was developed using machine learning algorithm using 3 classification models, which are KNN, Decision Tree, and Support Vector Machine. We used accuracy as the performance metrics to choose which model is the best. The result was that for graduation rate, the best model is SVM model with prediction accuracy value of 0.70 and for students performance, the best model is KNN with prediction accuracy value of 0.51.
Copyrights © 2023