Perbandingan Algoritma K Means dan K Medoids Untuk Clustering Kelas Siswa Tunagrahita

Harahap, Fitriana

TIN: TERAPAN INFORMATIKA NUSANTARA
Vol 2 No 4 (2021): September 2021

So far, the class placement of mentally retarded students is based on the child's entry age when registering at SLB C Muzdalifah, did not try the Intelligence Quotient (IQ) test for mentally retarded students in classifying student classes. This study aims to compare the results of class clustering for mentally retarded students using the K-Means and K-Medoids Clustering methods. The clusters produced by the two methods are 3. With the K-Means Clustering method, there are 8 students with mild mental retardation, 14 students with moderate mental retardation, and 14 students with severe mental retardation. Meanwhile, with the K-Medoids Clustering method, it can be seen that there are 7 students with mild mental retardation, 19 students with moderate mental retardation, and 10 students with severe mental retardation. The DBI value for K-Means validation is 0.161 and the DBI value for K-Medoids validation is 0.281. Thus, clustering using the K-Means Clustering method has better results than the K-Medoids Clustering method, because it produces a smaller DBI value of 0.161