Tuition Single Classification using Decision Tree Method and C4.5

Abstract

Tuition Single (UKT in Indonesia) is an operational cost measurement standard for universities in Indonesia which is applied by the government since 2013. This standard is enforced for every public university students who pass entrance university test. The problem with this standard is still found that there are students with tuition single category which does not fit with their condition, thus resulting in the incriminating of students with difficult economic situations. The usage of the decision tree and C4.5 algorithm for classification is able to process continuous data and build model efficiently with a large number of data. The input variables for tuition single classification are the income of the financier, the status of the children within the family, the taxable value of land and building, and the ownership of the vehicle. This work uses a multiclass variable. The output variables are group 1 to group 7. The accuracy of the system is 80.25% where the groups with the most students are group 3 and group 7. The decision tree built is also quite big considering the number of the variables and the variation of data used in each variable.