Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Dengue Risk Stratification in Semarang City Using a Gaussian Mixture Model Based on Multidimensional Urban Indicators

Rektorat UDINUS 2026 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2026
Goals
Record
8a65173a-a6d8-422f-84dc-fdc3f6d3225f
Published
2026-10-30

Universitas Dian Nuswantoro, through researchers Ismah and Nabila Izzatil Fahmi and Amiq Fahmi, conducted research on dengue risk stratification in Semarang City. The activity addressed dengue fever as a public health challenge in major Indonesian cities and examined how heterogeneous demographic structures and built-environment characteristics generate spatially uneven transmission risks. It also responded to limitations in conventional risk-mapping approaches, which may not capture the probabilistic nature of risk at fine-scale administrative levels.

The research developed a probabilistic and replicable framework using Gaussian Mixture Models. The method integrated multi-year dengue incidence data from 2021 to 2024 with eighteen multidimensional urban indicators across 177 sub-districts, referred to in the source as kelurahan. Unlike rigid partitioning methods, the approach accommodated overlapping data distributions and probabilistic cluster memberships. Several Gaussian Mixture Model configurations were evaluated using the Bayesian Information Criterion to identify the optimal number of clusters. The Bayesian Information Criterion declined markedly when the number of clusters increased from two to three, while further increases produced only marginal gains. The lowest value was achieved with three clusters, identified as the most parsimonious and effective solution.

Internal validation confirmed that the cluster structure captured epidemiological variance despite the heterogeneity of urban spatial data. Cluster 2 emerged as a geographically limited high-risk epicenter, with consistently elevated incidence, pronounced temporal variability, and extreme values. The findings indicate that dengue risk in Semarang is concentrated in localized areas of heightened vulnerability rather than being uniformly distributed. The framework is presented as replicable in other complex tropical urban environments and supports evidence-based public health decision-making and hyper-local intervention planning. Through this activity, UDINUS contributes research addressing health challenges and the analysis of urban conditions relevant to sustainable cities and communities.

Reference https://www.scopus.com/pages/publications/105029187366?origin=resultslist