Random Forest Risk Mapping of Foot-and-Mouth Disease in Sukabumi Regency, Indonesia

¹ Postgraduate Student of Veterinary Science Master Program, Faculty of Veterinary Medicine, Gadjah Mada University, Yogyakarta, Indonesia, Postal Code: 55281, Phone: +62 (274) 6492088 or +62 (274) 560862, Fax: +62 (274) 560861, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
² Sukabumi Regency Livestock Agency, West Java, Indonesia, Phone: +62 (266)533137, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
³ Department of Veterinary Public Health, Faculty of Veterinary Medicine, Gadjah Mada University, Yogyakarta, Indonesia, Postal Code: 55281, Phone: +62 (274) 6492088 or +62 (274) 560862, Fax: +62 (274) 560861, e-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.
⁴ Faculty of Veterinary Medicine Chiang Mai University 50100, Thailand
⁵ Department of Statistics, Faculty of Mathematics and Natural Science, Islamic University of Indonesia, Yogyakarta, Indonesia, Postal Code: 55584, Phone: +62 (274) 898444

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Foot-and-Mouth Disease (FMD) is a highly contagious viral disease that severely impacts livestock economies, especially in regions like Sukabumi Regency, Indonesia, where livestock farming is vital. This study aimed to assess the spatial risk of FMD in Sukabumi using Random Forest (RF) and Geographic Information Systems (GIS) to produce a comprehensive risk map. The objective was to identify high-risk zones and understand the key factors influencing FMD outbreaks in the region. The study employed an observational spatial epidemiological design, integrating historical outbreak data with 16 spatial risk factors, including livestock demographics, market and movement, veterinary capacity, and environmental variables. A Random Forest model was trained and validated using these predictors, with data split into training and testing subsets. Performance metrics, such as accuracy, sensitivity, specificity, and AUC, were used to evaluate model effectiveness. The results revealed that RF accurately classified FMD risk zones, with an AUC of 0.939, indicating high predictive performance. This study contributes to the understanding of FMD transmission dynamics and provides a practical framework for targeted disease control. This work demonstrates the potential of machine learning and GIS in improving livestock disease management strategies.