Clustering of Deleted Binjai City Government Asset Data Using the K-Means Algorithm

Regional assets are a crucial component in managing local government resources. However, their management often encounters various obastacles, such as the accumulation of unproductive assets and the difficulty of mapping assets that must be written off. The Binjai City Government, through the Regional Finance, Revenue, and Asset Management Agency (BPKPAD), is obliigated to manage its assets, including those that have reached the end of their useful life. However, without in-depth analysis, the management of written-off asset data can become disorganized, potentially hampering the transparency and efficiency of overall asset management. To address these issuses, this study applied the K-Means algorithm with 3, 4, and 5 clusters as a method for grouping deleted asset data. The data characteristics used included the type of item, year of acquisition, and method of acquisition. The test results showed that grouping with 3 clusters resulted in a cluster variance value of 208,6587, indicating a high level of data diversity. With 4 clusters, the cluster variance value decreased to 110,5156, resulting in a better and more compact grouping. Meanwhile, testing with 5 clusters provided the most optimal results, with a cluster variance value of 79,2477. This shows that the use of 5 clusters can minimize the spread of data within each cluster, resulting in higher similarity between data compared to 3 and 4 clusters. Therefore, the application of the K-Means Algorithm to deleted Binjai City Government asset data can assist the data analysis and grouping process, where the best results were obtained in testing with 5 clusters. Keywords: Asset Data, K-Means Clustering, MATLAB, Binjai City Government