PERBANDINGAN ALGORITMA C4.5 DAN ID3 UNTUK PREDIKSI KETEPATAN WAKTU LULUS MAHASISWA

Abstract

Student graduation data at a university, especially a study program, plays a very important role in evaluating the success of the educational program given to students. The education program is said to be good if the accuracy of student graduation is also good. In connection with the effort to evaluate educational programs in a study program, a study is needed to explore information from student graduation patterns with the Knowledge Discovering Database (KDD) data mining technique which includes the stages of data selection, preprocessing, data transformation, mining processes, and evaluation. The mining process uses the ID3 and C45 algorithms which are used to build a decision tree model. Based on the results of the implementation of the ID3 and C45 algorithms on student graduation testing data using RStudio, the highest accuracy value obtained by the C4.5 algorithm is 81.88% on the 30% testing data. Overall the results of the implementation of the data composition of 90%, 80%, 70%, 30%, 20%, and 10% obtained information that the accuracy value generated using the C4.5 algorithm is on average higher than the accuracy value of the ID3 algorithm. Thus it can be concluded that the best algorithm to predict the accuracy of student graduation is the C45 algorithm.

Keywords— ID3 Algorithm, C45 Algorithm, Accuracy, Prediction