PENENTUAN POLA MASA STUDI MAHASISWA JURUSAN MATEMATIKA DENGAN TEKNIK DATA MINING ALGORITMA C4.5 DI FAKULTAS MATEMATIKA DAN ILMU PENGETAHUAN ALAM UNIVERSITAS SRIWIJAYA
Daftar Isi:
- The students pattern of the study lenght can facilitate the stakeholders to improve education services in their institutions. This study investigated the graduated students pattern of the study from January 2014 until February 2018 with the C4.5 algorithm. Obtained data from three academic years 2011, 20112, and 2013 is 140 students an 18 frequently repeated study. The total attributs that used in the study is 19 attribute. These attributs containns of 18 frequently repeated subject and 1 is from student study lenght. This study used the method of K-fold cross-validation with K = 10 and the percentage split with the percentage is 66% to measure the performance of the algorithm C4.5. The results of both methods showed that the subjects that affect sequentially in the students’ study period are a complex function, numerical methods, introduction to computer science, computer programs and basic physics.The accuracy of the results obtained from both the training process K-fold cross validation amounted to 84,2857% and 72,9167% percentage split of algorithm C4.5 it shows that algorithm C4.5 is quite well in predicting the pattern of the study period students majoring in mathematics, State University of Sriwijaya.