PENDEKATAN GEOGRAPHICALLY WEIGHTED ZERO INFLATED POISSON REGRESSION (GWZIPR) DENGAN PEMBOBOT FIXED BISQUARE KERNEL PADA KASUS DIFTERI DI INDONESIA

Approach of Geographically Weighted Zero Inflated Poisson Regression (GWZIPR) With Weighting Function Used is Fixed Bisquare Kernel Case of Diphtheria in Indonesia

  • Ismah Ismah Universitas Muhammadiyah Jakarta
  • I Made Sumertajaya Prodi Statistika FMIPA IPB
  • Anik Djuraidah Pendidikan Matematika, Fakultas Ilmu Pendidikan, Universitas Muhammadiyah Jakarta
  • Anwar Fitrianto Prodi Statistika FMIPA IPB
Keywords: Geographically weighted zero inflated poisson regression,, fixed bisquare kernel

Abstract

The number of deaths due to diphtheria is counts data and there is a considerable presence of zeros (excess zeros). Besides, data on the spread of disease are generally geographically oriented or observed in each particular region, which is a type of spatial data. Geographically Weighted Zero Inflated Poisson Regression (GWZIPR), as the development of Geographically Weighted Regression (GWR) and Zero Inflated Poisson (ZIP) models will be used as a model in processing provincial diphtheria data in Indonesia in 2018, with the independent variable percentage of diphtheria cases (X1), percentage of vaccinated numbers (X2) and percentage of the population (X3) in each province in Indonesia. Estimating model parameters uses the method of maximum likelihood estimation. While the weighting function used is fixed bisquare kernel. Data is processed using software R packages lctools. The results were obtained if the model involved all three independent variables, the effect of the three independent variables on the number of deaths due to diphtheria was not significant. This is because there is a strong and significant relationship between independent variables, so that if the model does not involve a variable percentage of the population (population density), the percentage of vaccinated people affects the number of deaths caused by diphtheria significantly in an area. So that the provision of immunization vaccines can reduce the number of deaths caused by diphtheria

Downloads

Download data is not yet available.
Published
2020-03-01
How to Cite
[1]
I. Ismah, I. M. Sumertajaya, A. Djuraidah, and A. Fitrianto, “PENDEKATAN GEOGRAPHICALLY WEIGHTED ZERO INFLATED POISSON REGRESSION (GWZIPR) DENGAN PEMBOBOT FIXED BISQUARE KERNEL PADA KASUS DIFTERI DI INDONESIA”, BAREKENG, vol. 14, no. 1, pp. 039-046, Mar. 2020.