EARLY DETECTION MODEL OF DENGUE SPREADING AND ITS RELATION WITH CLIMATE AND TWOTTER DATA

• Main Author: NURUL FATIMAH (NIM: 10114016), SYIFA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/31203
• Institution: Institut Teknologi Bandung

Dengue Hemorrhagic Fever is a disease transmitted through an intermediary of <br /> <br /> mosquitoes. The number of dengue fever in Indonesia is the highest in ASEAN over other endemic areas in 2010. In 2016 dengue fever is included in the Extraordinary Occurrence with one of the factors is climate change. Therefore, making the early detection model of Dengue distribution is considered necessary, especially in areas that contribute a lot of dengue cases, for example in DKI Jakarta. Poisson, negative binomial, and logistic regression model, and deterministic model, that is SIR model will be used to know the relationship of DHF incidence with climate factor and twitter data. There are two types of units in a negative binomial regression model, weeks and months. Based on the simulation results, it was found that the model with unit of week with predictors of DHF data one and two weeks earlier and humidity one week before is the best model based on AIC method. While for model with month unit, it was found that the model with predictors of DHF data one and two months before and humidity one month before is the best model. Besides, the humidity threshold values for each Jakarta area that results in DHF incidence above the annual average is also found. The focus on the deterministic model is to find the parameter values in the model by using MATLAB. Based on the simulation result, it is found that the model with the parameter of a function has better result than the parameter which has constant value. In addition, the results of DHF incidence continues to decline, which means the longer time the DHF disease will be lost. <br />