Simulation Of The K-Means Clustering Algorithm With The Elbow Method in Making Clusters Of Provincial Poverty Levels in Indonesia

Abstract

One way to ensure that government programs and assistance for each province are right on target is to create a model of grouping or clustering provinces in Indonesia based on poverty levels. Algorithm K Means is one of the clustering methods in Data Mining to divide n observations into k groups so that each observation is in the group with the closest mean. In this study, provincial poverty level clustering in Indonesia will be made based on three poverty level indicators, namely the Percentage of Poor Population (P0), Poverty Depth (P1), and Poverty Severity (P2) with the K-Means Algorithm using the Elbow Method assisted by the Python Program. The results obtained are 5 optimal clusters of provincial poverty rates in Indonesia.