ANALISIS PREDIKSI KELULUSAN MAHASISWAFILKOM UNIVERSITAS NURDIN HAMZAH DENGAN METODE NAIVE BAYES

Abstract - Universities have an obligation to produce competent graduates. This can be measured from the graduation rate of students. The purpose of the analysis carried out is to predict graduation on time and find out indicators of problems that cause graduation not to be on time. Many factors influence students' late graduation, such as the student's marital status, the status of the student working or not working, the level of student understanding of the course material which can be seen from the student's GPA. Predicting student graduation is very important because this research can make a significant contribution both academically. Understand patterns or factors that influence student success. Provides insights for designing more effective educational interventions, such as academic tutoring. Based on the existing problems, it is necessary to have a system to predict student graduation rates based on existing variables. Attributes used are Name, Name, Gender, School of Origin, Study Program, Year of College Entry, Year of Graduation, Student Status, Marital Status, Age, IPS, GPA and Parent's Occupation. The method used is the Naive Bayes method. Comparison results were carried out using Excel and Rapidminer tools. The resulting accuracy level is 97,96% from 330 training data and 98 test data