Human Development Index Classification in Central Java Using the K-Nearest Neighbors Method for Data-Driven Policy Making

Human development in Central Java continues to show positive progress, as reflected in the consistent increase of the Human Development Index (HDI) across the province. The HDI serves as a key indicator for assessing the success of initiatives aimed at improving the overall quality of life. It measures how well residents are able to access the benefits of development, including long and healthy lives, education, knowledge, and a decent standard of living. The HDI is influenced by four primary components: life expectancy, expected years of schooling, mean years of schooling, and per capita expenditure. Currently, the Central Bureau of Statistics determines HDI values for each regency and city in Central Java using a specific calculation formula. In this study, we aim to classify these regions into three categories based on their HDI levels: very high, high, and moderate estimate areas. To perform this classification, we applied the K-Nearest Neighbors (KNN) algorithm—an effective, non-parametric method that classifies data points based on the majority class among their nearest neighbors in the feature space. KNN is well-suited for classification tasks involving complex, real-world data, offering both accuracy and interpretability. The classification of the 2024 HDI data using KNN resulted in three distinct groups: Cluster 1 (moderate estimate) includes 18 regions: Cilacap, Banyumas, Purbalingga, Banjarnegara, Kebumen, Wonosobo, Magelang, Wonogiri, Grobogan, Blora, Rembang, Temanggung, Kendal, Batang, Pekalongan, Pemalang, Tegal, and Brebes. Cluster 2 (high estimate) consists of 13 regions: Purworejo, Boyolali, Klaten, Sukoharjo, Karanganyar, Sragen, Pati, Kudus, Jepara, Demak, Semarang Regency, Kota Pekalongan, and Kota Tegal. Cluster 3 (very high estimate) comprises 4 urban areas: Kota Magelang, Kota Surakarta, Kota Salatiga, and Kota Semarang. This classification provides valuable insights into regional development disparities and can support evidence-based planning and policy-making.