Analysis of Hotels Spatial Clustering in Bali: Density-Based Spatial Clustering of Application Noise (DBSCAN) Algorithm Approach

Abstract

Bali is one of the hearts of tourism in Indonesia. The existence of the Covid-19 pandemic has made this tourist paradise also affected the wheels of the economy. Based on this, this study aims to determine the density clustering of one of the economic supporters in Bali, namely hospitality. The study began with the quadrant method and Ripley's K-Function to measure the distribution pattern of hospitality. From the results of the two methods, the distribution pattern of hotels in Bali is more towards clusters than random or regular distribution. If the point distribution pattern is more towards the cluster, it is continued with the Density-Based Spatial Clustering of Application Noise (DBSCAN) algorithm to form spatial clustering. In the DBSCAN algorithm, a combination of parameters, namely minimum points (MinPts) and epsilon (Eps), is carried out with evaluation using the silhouette average width value. From the results of the DBSCAN algorithm, the clustering results show that the distribution of hotels in Bali forms clusters and tends to approach the surrounding tourist attractions, such as near the beach, city market, and mountainous areas. It can help policymakers if they want to prioritize economic recovery after the Covid-19 pandemic.