Pengelompokan Kabupaten Dan Kota Di Provinsi Jawa Timur Berdasarkan Tingkat Kesejahteraan Dengan Metode K-Means Dan Density-Based Spatial Clustering Of Applications With Noise

Abstract

East Java Province has an uneven welfare condition. The uneven welfare conditions are indicated by a large number of poor people in East Java and the rate of economic growth which has decreased in 2020, reaching -2.39% due to the impact of the pandemic. Welfare can be measured through several indicators, while the indicators used to classify districts and cities in East Java among others include population density, labor force, labor force participation rate, and open unemployment rate. Thus, to find out the grouping of regencies and cities in East Java Province based on the level of welfare, grouping was carried out using the K-Means and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) methods. For each of the two methods, distance calculations are performed using the Euclidean and Manhattan distances. Each distance was tested for validity using the Davies-Bouldin Index (DBI), C-Index, and Dunn Index. This study concludes that the best method is the DBSCAN method using Manhattan distance with MinPts = 2 and eps = 4 which has the smallest DBI value of 0.284, with 2 clusters formed and 5 noise. Cluster 1 consists of 26 regencies, cluster 2 consists of 7 cities, and noise consist of 5 regencies and cities. Keywords: Welfare, K-Means, DBSCAN, Euclidean Distance, Manhattan Distance.