MODEL ALGORITMA KNN UNTUK PREDIKSI KELULUSAN MAHASISWA STIKOM CKI

Tiara Ratu Alifia (Sekolah Tinggi Ilmu Komputer Cipta Karya Informatika (STIKOM CKI))
Tundo Tundo (Sekolah Tinggi Ilmu Komputer Cipta Karya Informatika (STIKOM CKI))
Muhammad Syazidan (Sekolah Tinggi Ilmu Komputer Cipta Karya Informatika (STIKOM CKI))
Faldo Satria (Sekolah Tinggi Ilmu Komputer Cipta Karya Informatika (STIKOM CKI))

Jurnal Ilmiah Informatika Komputer
Vol 29, No 2 (2024)

This study develops a student graduation prediction model using the K-Nearest Neighbor (KNN) algorithm, considering variables such as age, Grade Point Average (GPA), number of Credits Earned (CE), participation in TOEFL tests, seminar activities, and participation in internships. Data from 80 students in the computer engineering and information systems programs at STIKOM Cipta Karya Informatika were analyzed to train and test the model. The results show that the KNN model with K=3, K=4, and K=5 produces a prediction accuracy of 66,67%. GPA and the number of credits earned significantly influence graduation, while participation in internships and TOEFL tests also contribute. Seminar certificates and age have a lower impact. These findings indicate that the KNN algorithm is effective for predicting student graduation, providing insights for educational institutions to enhance academic programs and student development.