Modeling of Dengue Hemorrhagic Fever Cases in West Java Using Gaussian, Poisson, and Compound Poisson-Tweedie Regression

Dengue Hemorrhagic Fever (DHF) is an infectious disease that remains a serious public health problem in Indonesia, particularly in West Java Province. This study aims to model the number of DHF cases using Gaussian, Poisson, and Compound Poisson Tweedie regression methods based on 2024 data sourced from the West Java Central Statistics Agency (BPS) and the West Java Provincial Health Office. The predictor variables used include humidity, rainfall, real per capita expenditure, poverty ratio, population density, percentage of households with clean and healthy lifestyles, health facility ratio, and average altitude above sea level. The analysis results show that Gaussian and Poisson regressions are less suitable because they cannot handle overdispersion in the data. The Compound Poisson Tweedie model provides the best performance with the smallest Akaike Information Criterion (AIC) value. Two variables that significantly influence the number of DHF cases are population density and the ratio of health facilities. The model equation can be expressed as follows: the expected number of DHF cases equals the exponential of (18.380 plus 0.00008 times population density minus 0.285 times health facility ratio).