Analisis Cluster Tingkat Perceraian Di Kota Malang Menggunakan Algoritma Agglomerative Hierarchical Clustering (Ahc)

Abstract

The high divorce rate in Malang City is a social issue that requires data-driven analysis to support targeted policy formulation, leading this study to analyze the patterns and characteristics of divorce at the Malang City Religious Court using the Agglomerative Hierarchical Clustering (AHC) algorithm with the Average Linkage method. Utilizing a dataset of 5,891 cases from the 2022–2024 period—including attributes such as divorce type, ages of the parties, causal factors, and the district of marriage—the research followed the CRISP-DM framework and employed the Davies-Bouldin Index (DBI) to determine that the optimal number of clusters was 10 for 2022, 2023, and the combined data, and 11 for 2024. The findings reveal that the 20–30 age range is the group most vulnerable to divorce, primarily driven by constant disputes and arguments, with Kedungkandang, Sukun, and Blimbing identified as the districts with the highest divorce rates. Ultimately, this research serves as a foundation for relevant authorities to design more targeted divorce prevention programs based on specific regional and demographic characteristics.