K-means cluster algorithm for grouping inequality in regional development

Unsupervised learning is a subset of machine learning. Many unsupervised learning algorithms are used to solve various problems, especially the extraction of hidden data patterns. One of the problems that concerns unsupervised tasks is clustering. Clustering techniques are widely used for data grouping needs, one of which is development inequality clustering. The classification of development inequality is an important consideration in a country's regional development strategy. However, development groupings often do not pay attention to the hidden information aspects of the data, so they do not get the appropriate results. This research was conducted to provide an additional alternative in the realm of development inequality clustering, namely by classifying development data using the k-means algorithm. The data used is GRDP data for 41 regions in the western part of Java Island for the 2010–2021 range. The results show that the forty-one regions are grouped into four clusters. The first cluster (C1) contains 35 regions, the second cluster (C2) contains three regions, the third cluster (C3) contains four regions, and the fourth cluster (C4) contains three regions. Based on the cluster results, it can be seen that all members of cluster C4 are areas with the best level of development, while cluster C1 is the area with the lowest level of development. As for clusters C2 and C3, these are areas with development levels between clusters C1 and C4. The grouping results can be used by policymakers or local governments to determine the direction of future development priorities based on the cluster with the lowest level of development.