Algoritma K-Means Clustering Penggunaan Bandwidth Internet (Studi Kasus di Pemerintah Daerah Kabupaten Padang Pariaman)

Abstract

To support government activities, a fast and precise network connection is needed. So it requires a wide network bandwidth. Bandwidth management needs to be done so that the network speed remains stable. This study aims to look at the pattern of bandwidth usage in the Regional Government of Padang Pariaman Regency using K-Means Clustering. The data is taken from the Cacti application, an open-source, web-based network monitoring software. The total extracted datasets used are 32 OPD data (Regional Apparatus Organizations) in the Regional Government of Padang Pariaman Regency in 2022. The available data is then processed to obtain cluster targets by utilizing the data mining concept using the K-Mean Clustering method. Bandwidth usage data grouping in Padang Pariaman Regency uses the Clustering method with the K-Means algorithm with the attributes Name OPD, Inbound Average, Inbound Maximum, Outbound Average, Outbound Maximum used in the process of calculating and dividing data into 3 clusters with high bandwidth usage categories, low, and medium. Calculations are done manually and then tested with RapidMiner software. The results of the manual calculations obtained the same number of cluster members as the calculations with the RapidMiner software.