Prediksi Masa Studi Mahasiswa Unama Jambi Menggunakan Metode Algoritma C4.5

Abstract

Listen

Translate article

Every year the number of students at The University of Dynamics of the Nation jambi is always increasing but the students who graduate are different from the number of students who enter. Therefore, the author conducts a data mining analysis on student data so that it can be used by academic supervisors to find out the graduation status of students and as a warning so that students can graduate on time so as to reduce the number of delays in graduation. The author uses student data in 2016 and 2017 as training data and 2020 as testing data as many as 120 training data and 109 testing data and has carried out the data cleaning and attribute selection process using the forward selection method. In conducting the analysis, the author used the Tools Weka tool. The author uses the C4.5 algorithm method with 12 attributes, but there are 4 attributes that are most influential after selecting data using the forward selection method on WEKA. In this case, the author uses 4 test options, namely 5 Fold Cross Validation, 10 Fold Cross Validation, 70% Percentage Split, and 80% Percentage Split. The C4.5 Algorithm method produces the largest accuracy value in the training data, namely 5 Fold Cross Validation by 92.5% and in the testing data by 100%.