Open AccessDeteksi Dini Mahasiswa Drop Out Menggunakan C5.0
Author(s) -
Ulfi Saidata Aesyi,
Alfirna Rizqi Lahitani,
Taufaldisatya Wijatama Diwangkara,
Riyanto Tri Kurniawan
Publication year - 2021
Publication title -
jiska (jurnal informatika sunan kalijaga)
Language(s) - English
Resource type - Journals
eISSN - 2528-0074
pISSN - 2527-5836
DOI - 10.14421/jiska.2021.6.2.113-119
Subject(s) - drop out , attendance , decision tree , statistic , computer science , statistics , engineering , artificial intelligence , mathematics , political science , economics , law , demographic economics
The decline in the number of active students also occurred at the Faculty of Engineering and Information Technology, Universitas Jenderal Achmad Yani. This greatly affects the profile of study program graduates. So it is necessary to have a system that is able to detect students who are threatened with dropping out early. In this study, the attributes chosen were the student's GPA and the percentage of attendance . This attribute is used to classify students who are predicted to drop out. The research data uses student data from the Faculty of Engineering and Information Technology, Universitas Jenderal Achmad Yani. This study uses the C5.0 algorithm to build a decision tree to assist data classification. The decision tree that was built with 304 data as training data resulted a C5.0 decision tree which had an error rate of 5%. The accuracy results obtained from the 76 test data is 93%.