Implementasi Algoritma K-Nearest Neighbor Untuk Memprediksi Program Studi Bagi Calon Mahasiswa Baru

Abstract

One of the factors for students graduating on time with good grades is that the study program they take is in accordance with their interests and competencies. For this reason, in the process of admitting new students, it is necessary to carry out selection, information and direction regarding the chosen study program. By using previous year's student data, data mining processing is carried out to produce classifications of study programs for prospective new students. To get maximum results, preprocessing data is carried out, after which the data is divided into training data and testing data. The two data are then processed with the K-Nearest Neighbor algorithm to determine the suitability of the Study Program class in the testing data and then the measurement accuracy value is calculated. Because it has a high accuracy value of 74%, using this training data it is developed in the form of an application with Java NetBeans which can be used to assist prospective new students in predicting the appropriate study program.