PENERAPAN DATA MINING UNTUK PREDIKSI KELULUSAN MAHASISWA FAKULTAS TEKNIK INFORMATIKA UNIVERSITAS JABAL GHAFUR MENGGUNAKAN METODE K-NEAREST NEIGHBOR BERBASIS WEB

Abstract

One of the instruments in campus accreditation, especially the faculty of informatics engineering at Jabal Ghafur University in order to get a good grade score is the satisfaction of graduate students getting decent jobs according to their fields, and is also a consideration during the study period for completing long lectures in a study program or faculty. The prediction of student graduation graduating on time or late does not only look at student data based on the highest scores as has been done in conventional grading systems so far. With advances in computer technology, especially in the field of data mining, it has brought many changes in the process of analyzing data patterns with data mining techniques, for example, in determining student graduation, a K-Nearest Neighbor method is used which processes data mining based on test data and sample data, especially on students. The results of the process for predicting student graduation are used sample data and training data. The sample data are students who are currently undergoing lectures and training data, namely students who have become alumni based on their graduation parameters. The final result obtained in the thesis research is that the system can input alternative data, criterion data, and process graduation predictions using the K-nearest neighbor method so that it can make a decision on whether the student who is being tested is "Passed" or "Graduated Late". The system can also display the K-nearest neighbor manual calculation flow and can display results reports.

Key Words : Datamining, K-nearest Neighbor, Student Graduation Prediction, , PHP&MySQL, Jabal Ghafur University