Clustering sub-districts based on the number of disasters in East Java using the hierarchical cluster method
Abstract
Calamities of a natural origin are those precipitated by occurrences or sequences of phenomena instigated by the forces of nature, encompassing seismic disturbances, oceanic surges, volcanic outbursts, inundations, arid spells, cyclonic tempests, and terrestrial slippages. Natural disasters in East Java are one of the concerns of the government and the community. Throughout 2022 to 2023 in East Java, various natural disasters were recorded, including floods, earthquakes and landslides. This study focuses on clustering sub-districts in East Java based on the number of disasters that occurred, namely floods, earthquakes and landslides, using four hierarchical clustering methods: single linkage, complete linkage, average linkage, and ward. In comparison, it was found that the ward method is the most optimal method based on the icdrate value. Based on the outcomes of the optimal cluster graph and the lowest icdrate value, 5 optimal clusters were obtained from the grouping of natural disaster events that occurred in each sub-district. This finding is important for disaster mitigation planning in line with SDGS 13 "Climate Change Management" by emphasizing the effectiveness of the ward method in identifying disaster risk patterns for the development of more appropriate mitigation strategies.
Fields of Study & Subjects
Data Mining and Machine Learning Applications
Disaster Management and Resilience
Advanced Clustering Algorithms Research
Publication details
- Journal / source
- AIP conference proceedings
- Volume
- 3326
- Pages
- 080028–080028
- Publication date
- 1 January 2026
- Language
- EN