Comparison of K-Means and K-Medoid Algorithms in Classifying Village Status (Case Study: Gorontalo Province)

Abstract

In the national development process, the village occupies a very important position. This is because it is the smallest government structure and has direct contact with the community. Seeing the importance of its role in national development, one of which is Gorontalo Province, based on directions from the central government, is trying to implement the Village Fund Allocation (ADD) policy for all villages in Gorontalo Province. In distributing the allocation of funds, it is necessary to map the status of the Village to find out the amount that must be given. This test uses the average execution time and the Davies Bouldin Index (DBI). After testing it is known that the K-Medoid Algorithm has a better DBI value than the K-Means Algorithm with the DBI value of the K-Medoid Algorithm being 0.050. On the other hand, the K-Means Algorithm has a better average execution time than the K-Medoid Algorithm, where the average execution time is 1 second.