Uang Kuliah Tunggal (UKT) is a portion of the single tuition fee that is borne by each undergraduate student at a state university in Indonesia. UKT is the number of fees that have to be paid by students in each semester. Basically, UKT is implemented to impose tuition fees according to the income and the condition of students’ families. However, there is an issue in regard to the inappropriate classification of UKT. This issue is caused by several factors such as manual method is still used in determining UKT Classes and there is likely an element of subjectivity in determining UKT Classes of new students. Based on these issues, a decision support system that can determine UKT Class of new students is needed. The Naïve Bayes Classifier (NBC) method is used to classify data into eight UKT Classes. Whereas Fuzzy-TOPSIS is used in the optimization selection process of UKT Class 1 to 8. This method was chosen for its capability in choosing the best alternative out of several possibilities, in this case, the intended alternative is the most suitable new student to be selected in UKT Class based on six predetermined criteria. Research results show that the NBC Model can identify UKT groups of students with mean values of precision and recall testing are 77.8% and 77.8% and the model accuracy is 77.8% as well. In optimization selection process of UKT using Fuzzy-TOPSIS results obtained the percentage of UKT 1 group recommendations was 5.33%, UKT 2 was 5.33%, UKT 3 was 10.22%, UKT 4 was 24.89%, UKT 5 was 24.89%, UKT 6 was 10.22%, UKT 7 was 10.22% and UKT 8 was 8.00%. Based on the results above, it can be concluded that the combination of Naïve Bayes Classifier and Fuzzy-TOPSIS can be implemented for determining the UKT Classes. Then the recommendations of the UKT Classes can be considered in determining the UKT Classes of students for the Decision-maker.

No abstract