PREDIKSI KELULUSAN TEPAT WAKTU MAHASISWA TEKNIK ELEKTRO UMM MENGGUNAKAN METODE NAÏVE BAYES DAN KNN

At the end of this end it was discovered that each year accepted new students, but this was not offset by the graduation rate of the students, where only a few students could graduate on time. therefore, students graduate during the time of the University's provisions, which can support progress and increase accreditation. The study used naive bayes and KNN techniques. In the Naïve Bayes Algorithm, a statistical classification process is carried out to predict the probability of a class. Meanwhile, the K-Nearest Neighbor (K-NN) algorithm is carried out a classification process for new objects based on (K) their closest neighbors. The results of the study showed that the accuracy rate of the naïve bayes algorithm was 42% while the K-NN algorithm was 65%. That way it can be concluded that the K-NN algorithm has more accuracy than the naïve bayes algorithm in predicting student graduation