PREDICTING STUDENT GRADE POINT AVERAGE WHO IS STUDYING WHILE WORKING AT ADVENTIST UNIVERSITY OF INDONESIA USING DECISION TREE C4.5 METHOD AND SMOTE
Keywords:
Data Mining, Decision Tree C4.5, SMOTE, Predicting Student GPA, Studying While WorkingAbstract
Higher education is one way to get job easier, this thing happens because through education the individual is able to increase the level of human resources in this era. However, the high cost of education is very expensive so individuals who wants to study must also work at the same time, so this research aims to predict the student GPA who is studying while working at the same time at Adventist University of Indonesia. From the results of this research there are 8 attributes that have an effect on predicting student GPA at Adventist University of Indonesia, namely the Department of Work, Working Hours, Course, Gender, Residence, Age and Number of Credits. The method that has been used in this research is Decision Tree C4.5 implemented on the WEKA program with the J48 algorithm. This research also uses the SMOTE (Synthetic Minority Oversampling Technique) algorithm to balancing the amount of data in the minor class. The top root of this research is Gender which affects the student GPA at University of Indonesia. The SMOTE algorithm in this research is useful to help raising the result of this research by 7-8% can be seen from the results of the accuracy of the cross validation 10 folds test is 63.6672%, the average result of precision and recall are 0.621 and 0.637. While the accuracy of the split test 70:30 is 62.7955%, then result of precision and recall are 0.621 and 0.628. When compared with the use of the Decision Tree C4.5 algorithm only, the accuracy of the cross validation 10 fold test is 55.5044%, with the average result of precision and recall is -.545 and 0.555. While the accuracy of the split test 70:30 is 55.2995% with the results of precision and recall is 0.554 and 0.553. The analysis results using confusion matrix and ROC curve with results from 0.688 to 0.756, which are in the range of 0.70 - 0.80 which is included in the level of fair classification diagnosis. It can be concluded that there is a strong effect while working on the student GPA. With the order of attributes from the top most are Gender, Total Credit, Department, Age, Department of Work, Working Hours and Residence.
Downloads
References
Andriyan, David. (2016). Indeks Prestasi Komulatif Mahasiswa Ditinjau dari Strategi Belajar dan Keaktifan Berorganisasi pada Mahasiwa Pendidikan Akuntansi Universitas Muhammadiyah Surakarta tahun 2014. [Online]. Available: http://davidandriyan.blogspot.co.id/2016/07/proposal-penelitian-dengan-judul-indeks.html [5 April 2018]
B. Rossi, S. Itasia, and A, Farit, “Penerapan Synthetic Minority Oversampling Technique (SMOTE) Terhadap Data Tidak Seimbang Pada Pembuatan Model Komposisi Jamu,” vol. 1, no. 1, 2013. Diakses pada: Maret, 16, 2019. [Online]. Tersedia di: http://dx.doi.org/10.29244/xplore.v1i1.12424
Berry, Michael J.A dan Linoff, Gordon S (2004). Data Mining Techniques For Maketing, Sales, Customer Relationship Management Second Edition. United States of America: Wiley Publishing, Inc.
Defiyanti, Sofi (2014). Perbandingan: Prediksi Prestasi Belajar Mahasiswa Menggunakan Teknik Data Mining (Study Kasus Fasilkom UNSIKA). Makasar: KNSI.
Larose, D. T. (2005). Discovering Knowledge in Data. New Jersey: John Willey &Sons, Inc.
Lestari, Fenti (2016). Pengaruh Lingkungan Keluarga Dan Fasilitas Belajar Terhadap Motivasi Belajar Dan Hasil Belajar Siswa KElas XI IPS Pada Mata Pelajaran Ekonomi DI SMAN 2 Kebumen Tahun Pelajaran 2015/2016. Yogyakarta: Universitas Negri Yogyakarta.
Putri, Ratna P.S dan Waspada, Indra (2018). Penerapan Algoritma C4.5 Pada Aplikasi Prediksi Kelulusan Mahasiswa Prodi Informatika. Semarang: Jurnal Ilmu Komputer dan Informatika
Downloads
Published
How to Cite
Issue
Section
License
The submitting author warrants that the submission is original and that she/he is the author of the submission together with the named co-authors; to the extend the submission incorporates text passages, figures, data or other material from the work of others, the submitting author has obtained any necessary permission.
Articles in this journal are published under the Creative Commons Share Alike Attribution Licence (CC-BY-SA What does this mean?). This is to get more legal certainty about what readers can do with published articles, and thus a wider dissemination and archiving, which in turn makes publishing with this journal more valuable for you, the authors.
By submitting an article the author grants to this journal the non-exclusive right to publish it. The author retains the copyright and the publishing rights for his article without any restrictions.
