Open access Jurnal SINTA 2
Prediksi Kelulusan Tepat Waktu Berdasarkan Riwayat Akademik Menggunakan Metode K-Nearest Neighbor
Jurnal Teknologi Informasi dan Ilmu Komputer (JTIIK) · Vol. 11 No. 2 (2024) · pp. 249–256
- ISSN
- 2528-6579, 2355-7699
- Published
- 2024-08-26
- Publisher
- Fakultas Ilmu Komputer, Universitas Brawijaya
- Language
- en
Abstract
Bahasa Indonesia
Lulus tepat waktu adalah pencapaian paripurna ingin dicapai oleh semua mahasiswa atau calon mahasiswa. Kelulusan memiliki 2 klasifikasi seperti lulus tidak tepat waktu dan lulus tepat waktu. Kelulusan menjadi suatu penilaian akreditasi universitas dan penilaian oleh masyarakat secara luas. Perguruan tinggi meluluskan mahasiswa-mahasiwa dengan beberapa kriteria standar yang harus dimiliki. Diharapkan mahasiswa lulus memenuhi syarat standar kelulusan dalam waktu maksimal 4 tahun masa studi. Evaluasi dan pemantauan kelulusan sangat penting dilakukan, salah satunya dengan mempelajari data history mahasiswa yang telah lulus sebagai upaya mahasiswa lulus tidak melebihi waktu standar yang telah ditetapkan. Prediksi kelulusan yang dilakukan pada riset menggunakan kaidah klasifikasi K-Nearest Neighbor dengan objek penelitian yaitu mahasiswa. Atribut yang dipakai dalam penelitian metode klasifikasi yaitu nama, asal SMA/SMK, wilayah asal SMA/SMK, nilai rata-rata matematika dan lama studi. Tahapan penelitian diawali dengan pengumpulan data, pemilihan atribut, pembersihan data, transformasi data, pemilihan data testing dan data training. Pengujian akurasi yang didapatkan pada penelitian metode klasifikasi dengan klaster data k=1, k=2, k=3, k=4, k=5, k=6 dan k=7 menghasilkan klaster dengan nilai k=3 paling tinggi. Hasil pengujian akurasi prediksi penelitian menggunakan confusion matrix menghasilkan akurasi paling besar sesuai target yaitu mencapai 78% menggunakan objek penelitian sebanyak 93 data mahasiswa terdiri dari 78 training dan 12 data testing. Hasil pengujian point k=1 sampai point k=7, k=3 merupakan nilai akurasi prediksi yang paling tinggi sehingga hasil penelitian menjadi sumber pengetahuan untuk fakultas dalam prediksi kelulusan mahasiswa.AbstractGraduating on time is a plenary achievement to be achieved by all students or prospective students. Graduation has 2 classifications such as graduating not on time and graduating on time. Graduation becomes an assessment of university accreditation and an assessment by the wider community. Universities graduate students with several standard criteria that must be possessed. It is expected that graduating students meet the graduation standard requirements within a maximum of 4 years of study period. Evaluation and monitoring of graduation is very important to do, one of which is by studying the history data of students who have graduated as an effort for students to graduate not to exceed the standard time that has been set. The graduation predictions carried out in research use the K-Nearest Neighbor classification rules with the research object being students. The attributes used in the research classification method are name, high school/vocational high school origin, high school/vocational high school origin, average grade in mathematics and length of study. The research phase begins with data collection, attribute selection, data cleaning, data transformation, selection of testing data and training data. The accuracy test obtained in the classification method research with data clusters k = 1, k = 2, k = 3, k = 4, k = 5, k = 6 and k = 7 produces a cluster with the highest k = 3 value. The results of testing the accuracy of research predictions using the confusion matrix produced the greatest accuracy according to the target, reaching 78% using a research object of 93 student data consisting of 78 training and 12 testing data. The test results point k=1 to point k=7, k=3 is the highest prediction accuracy value so that the research results become a source of knowledge for the faculty in predicting student graduation.
English
Abstract Graduating on time is a plenary achievement to be achieved by all students or prospective students. Graduation has 2 classifications such as graduating not on time and graduating on time. Graduation becomes an assessment of university accreditation and an assessment by the wider community. Universities graduate students with several standard criteria that must be possessed. It is expected that graduating students meet the graduation standard requirements within a maximum of 4 years of study period. Evaluation and monitoring of graduation is very important to do, one of which is by studying the history data of students who have graduated as an effort for students to graduate not to exceed the standard time that has been set. The graduation predictions carried out in research use the K-Nearest Neighbor classification rules with the research object being students. The attributes used in the research classification method are name, high school/vocational high school origin, high school/vocational high school origin, average grade in mathematics and length of study. The research phase begins with data collection, attribute selection, data cleaning, data transformation, selection of testing data and training data. The accuracy test obtained in the classification method research with data clusters k = 1, k = 2, k = 3, k = 4, k = 5, k = 6 and k = 7 produces a cluster with the highest k = 3 value. The results of testing the accuracy of research predictions using the confusion matrix produced the greatest accuracy according to the target, reaching 78% using a research object of 93 student data consisting of 78 training and 12 testing data. The test results point k=1 to point k=7, k=3 is the highest prediction accuracy value so that the research results become a source of knowledge for the faculty in predicting student graduation. Abstrak Lulus tepat waktu adalah pencapaian paripurna ingin dicapai oleh semua mahasiswa atau calon mahasiswa. Kelulusan memiliki 2 klasifikasi seperti lulus tidak tepat waktu dan lulus tepat waktu. Kelulusan menjadi suatu penilaian akreditasi universitas dan penilaian oleh masyarakat secara luas. Perguruan tinggi meluluskan mahasiswa-mahasiwa dengan beberapa kriteria standar yang harus dimiliki. Diharapkan mahasiswa lulus memenuhi syarat standar kelulusan dalam waktu maksimal 4 tahun masa studi. Evaluasi dan pemantauan kelulusan sangat penting dilakukan, salah satunya dengan mempelajari data history mahasiswa yang telah lulus sebagai upaya mahasiswa lulus tidak melebihi waktu standar yang telah ditetapkan. Prediksi kelulusan yang dilakukan pada riset menggunakan kaidah klasifikasi K-Nearest Neighbor dengan objek penelitian yaitu mahasiswa. Atribut yang dipakai dalam penelitian metode klasifikasi yaitu nama, asal SMA/SMK, wilayah asal SMA/SMK, nilai rata-rata matematika dan lama studi. Tahapan penelitian diawali dengan pengumpulan data, pemilihan atribut, pembersihan data, transformasi data, pemilihan data testing dan data training. Pengujian akurasi yang didapatkan pada penelitian metode klasifikasi dengan klaster data k=1, k=2, k=3, k=4, k=5, k=6 dan k=7 menghasilkan klaster dengan nilai k=3 paling tinggi. Hasil pengujian akurasi prediksi penelitian menggunakan confusion matrix menghasilkan akurasi paling besar sesuai target yaitu mencapai 78% menggunakan objek penelitian sebanyak 93 data mahasiswa terdiri dari 78 training dan 12 data testing. Hasil pengujian point k=1 sampai point k=7, k=3 merupakan nilai akurasi prediksi yang paling tinggi sehingga hasil penelitian menjadi sumber pengetahuan untuk fakultas dalam prediksi kelulusan mahasiswa.
Collaboration network of the authors
22 authors · 46 co-authorship links · 3 across different organisations
Hover or tap a name to highlight its co-authors. Double-click to open the author's page.
- Universitas Ahmad Dahlan
- Islamic University of Indonesia
- Affiliation unknown
- Different organisations
- Same organisation
- Affiliation unknown
- Thicker line = more joint articles · bigger circle = more articles
Show as table
| Co-author | Organisation | Joint articles |
|---|---|---|
| Anton Yudhana | Universitas Ahmad Dahlan | 5 |
| Abdul Fadlil | Universitas Ahmad Dahlan | 3 |
| Sunardi Sunardi | Universitas Ahmad Dahlan | 2 |
| Yudi Prayudi | Islamic University of Indonesia | 2 |
| Siti Helmiyah | Universitas Ahmad Dahlan | 2 |
| Abdullah Hanif | Universitas Ahmad Dahlan | 2 |
| Ainunna’imah | Universitas Ahmad Dahlan | 1 |
| Herman Yuliansyah | Universitas Ahmad Dahlan | 1 |
| Herman | Universitas Ahmad Dahlan | 1 |
| Fitriah | Universitas Ahmad Dahlan | 1 |
| Asno Azzawagama Firdaus | Universitas Ahmad Dahlan | 1 |
| Yuminah yuminah | Universitas Ahmad Dahlan | 1 |
| Imam Mahfudl Nasrulloh | Universitas Ahmad Dahlan | 1 |
| Fauzan Natsir | Affiliation unknown | 1 |
| Eri Haryanto | Islamic University of Indonesia | 1 |
| Sahiruddin Sahiruddin | Universitas Ahmad Dahlan | 1 |
| W Yunanri | Universitas Ahmad Dahlan | 1 |
| Andi Prayudi | Universitas Ahmad Dahlan | 1 |
| Nurmi Hidayasari | Islamic University of Indonesia | 1 |
Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.