Work

Analysis of Geographical Characteristics and Risk Factors for Non-Communicable Diseases: Diabetes in Central Java Using Random Forest and SHAP

Year: 2025

Type: conference-paper

Abstract: Diabetes mellitus is a growing health issue in Indonesia, particularly in Central Java Province.In 2023, it contributed significantly to public health concerns due to its increasing prevalence.This study utilizes the Random Forest algorithm to identify dominant diabetes risk factors and map their geographic distribution across the region.The analysis incorporates community lifestyle indicators such as the number of sweet drink stores, gyms, sports halls, transportation availability, karaoke venues, tourist attractions, fast food outlets, and alcohol stores in each regency/city.Results indicate that the number of tourist attractions (feature importance: 0.2817) and sweet drink stores (feature importance: 0.1502) are the primary global risk factors.Locally, tourism dominates as the key risk factor in 18 regencies/cities (51.4% of the region).The model employs optimal hyperparameter tuning to improve accuracy and utilizes SHAP techniques to evaluate the importance of local features.This study aims to enhance understanding of the key risk factors contributing to diabetes cases and their distribution in Central Java Province.

Source: Proceedings of International Conference on Artificial Life and Robotics

DOI: 10.5954/icarob.2025.os26-6

Authors: Ambar Arum Prameswari, Andi Prademon Yunus

Institution: Telkom University

Language: English

Topic: Artificial Intelligence in Healthcare

Subfield: Health Information Management

Field: Health Professions

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: bronze

Funder: Universitas Telkom