The Implementation of Data Mining to Show UKT (Students’ Tuition) Using Fuzzy C-Means Algorithm : (Case Study: Universitas Pendidikan Ganesha)

This research aimed to show the result of clustering students' tuition (UKT) at Undiksha using algorithm FCM. The characteristics of each cluster, measurement of level implementing algorithm FCM accuracy in determining UKT. Students' tuition data used in this research include students' tuition from SBMPTN year 2017. The students' data came from 30 students with 7 parameters, namely, parents' occupation, parents' income, number of dependents, assets, water payment, electronic voltage, and varieties of vehicles. The data of students' tuition grouped into four groups, namely, UKT 1, UKT 2, UKT 3, and UKT 4. The data from grouping students' tuition using FCM method in determining students' tuition supported with Matlab Software 2017 a showed UKT 1 into 89 students, UKT 2 into 91 students, UKT 3 into 79 students, and UKT 4 into 46 students. The data characteristics of each student's tuition were gathered from each parameter based on the result of the center vector (v) in the last iteration. Besides, the result showed an FCM method has high accuracy in 0.78. The result of factor analysis showed 3 factors determined students' tuition from 7 parameters, namely, income factor, expulsion factor, and load factor. On the other hand, future research can be developed by grouping the 3 factors as computation variable in algorithm FCM and to use other methods, so that the results of clustering are more optimal.