Geographically weighted count modelingtechniques applied to dengue data analysis

• Main Authors: Hsin-Yu Chu, 屈欣諭
• Other Authors: 陳怡如
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/8nqh7s

碩士 === 淡江大學 === 統計學系應用統計學碩士班 === 106 === In recent years, the Dengue fever continued to spread all over the world. In 2015, the cases of Dengue fever exceeded 40,000. Most researchers focused the research in Tainan, Kaohsiung and Pingtung, where the epidemic is more serious, the rarely discussed the epidemic all over Taiwan. In this thesis, we focused on analyzing the Dengue fever throughout Taiwan in 2015, using Geographically Weighted Poisson Regression, Geographically Weighted Negative Binomial Regression, Geographically Weighted Zero-Inflated Poisson Regression, Geographically Weighted Zero-Inflated Negative Binomial Regression analysis to find out whether social factors and environmental factors have the spatial nonstationary effect on the cumulative case weeks. However, the Dengue fever is a disease that rarely occurs, causes no cases occurred in some counties and towns. Thus, the distribution of the Dengue fever is unbalanced and overdispersed. Therefore, our goal is to find out which model fits better. Moreover, by analyzing the cumulative case weeks and the impact factors, we expected to develop different prevention measures which can prevent Dengue fever from breaking out.