NEGATIVE BINOMIAL REGRESSION IN DYNAMIC MAPPING OF DENGUE FEVER INCIDENTS IN DKI JAKARTA WITH CLIMATE FACTORS

• Main Author: Febria Finola, Clarissa
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39212
• Institution: Institut Teknologi Bandung

DHF, referred to Dengue Hemorrhagic Fever, is an infectious disease with Aedes sp. as the infectious vector. This epidemic spreads mostly in the tropical region, and the other small portion in the subtropical region. Its incidence rate in Indonesia is quite high at the beginning of every year. Climate change is suspected to be one of the main causes. DHF, which is a count data, can usually be modeled as response data with Poisson or Binomial Negative regression. Therefore, the purpose of this final project is to find out how well the Poisson or Negative Binomial regression in modeling the incidence of DHF in DKI Jakarta by its relation with climate factors including air temperature, rainfall, and humidity, and also the incidence of DHF in the past. Before building the model, it was examined first which predictor variables were sufficiently correlated with the response variable. Then, the best model was selected based on the smallest AIC value, the smallest RMSE value, and the highest coefficient of determination. Temporal models obtained from each sub-district will be used in a spatial mapping of the spread of dengue incidence in each month. This final project is expected to provide additional insight to the community regarding the application of mathematics in the health field and especially for the Dinas Kesehatan DKI Jakarta, so that the future dengue events can be detected early and then the government immediately takes precautionary measures.