Determining Natural Disaster Mitigation Level using Unsupervised k-means Clustering
Abdurrakhman Prasetyadi, Budi Nugroho, Merios Gusan Putra · 2022
This work intends to categorize the number of districts/cities based on their mitigation efforts using unsupervised clustering approaches to generate decision support systems. Using data mining techniques and k-means clustering algorithms, it is possible to address problems involving data on the number of districts/cities based on natural disaster mitigation measures. Ms. Excel is used to estimate the value of the centroid for three clusters: the high expectation level cluster (C1), the medium anticipation level cluster (C2), and the low anticipation level cluster (C3) (C3). Our trials highlighted the outcomes of categorizing districts/cities based on their natural disaster preparedness efforts with two high-level districts/cities, namely Mentawai Islands Regency and South Lampung Regency, 47 regencies/cities at medium level 32 other districts/cities including low-level clusters. The results can be adopted as a recommendation for the district/city government to improve facilities and infrastructure in the district/city conforming to the efforts of natural disaster mitigation based on the clusters.