Pengelompokan Data Mahasiswa menggunakan Algoritma K-Means

Abstract

New Student Registration at Universitas Katolik De La Salle Manado (UKDLSM) is open every year starting in February. Most prospective students come from within the North Sulawesi province; many even come from outside the province, such as Gorontalo, South Sulawesi, Papua, and Maluku. This student data will later be useful and produce much information if processed properly. Through this data, you can see the distribution of students who enter this university from which areas, the type of high school or vocational school equivalent, and many more. Therefore, this research develops a data mining-based application to cluster student data at UKDLSM using the K-Means Clustering algorithm. The data parameters processed include the district/city of origin, study program, school of origin, and department at the school of origin. The clustering process is done by selecting the value K=3 according to student data of the 3 study programs with the highest number of students. It produces the amount of data in each cluster, namely cluster 1 with 408 data, 2 with 19 data, and cluster 3 with 58 data, with several iteration processes: 7 iterations and Silhouette Score: 0.56.