EVALUATION OF DATA MINING MODELS FOR STUDENT GRADUATION PREDICTION USING NAÏVE BAYES, KNN, AND C4.5 ALGORITHMS

This research was conducted to help colleges or universities monitor or predict student graduation. Admission selection is carried out to attract high-quality students. Student quality can be measured by the length of their stay at the university. Graduated on time with the title of Outstanding Student. The number of degrees has a significant impact on the university's accreditation ranking. Many factors, such as study program, admission route, gender, and Semester Achievement Index, have a significant impact on when a student can graduate in their studies. These variables are processed using the KNN (K-Nearest Neighbors) algorithm, Naïve Bayes, C4.5. Data preprocessing uses student data such as study program, gender, admission route, and semester Achievement Index from semesters 1 to 6. Based on graduation data, 452 students graduated on time and 48 students graduated late. All data is taken from the graduation database from 2018 to 2022.