Pemodelan Geographically Weighted Regression (Gwr) Dengan Fungsi Pembobot Adaptive Kernel Bisquare Untuk Angka Kesakitan Demam Berdarah di Kalimantan Timur Tahun 2015

Abstract

Dengue Fever in East Borneo is thought to be a spatial problem that affected by geographic factor and linear regression analysis that is often can not describe with Good Relations pattern. The solution for this problem can be solved using Geographic Weighted Regression Method (GWR) to review and Troubleshooting geographic factor. This research Model proposed to consider GWR model with geography factor or location as the weight to estimate the model parameters, the weight type that used for this research is Adaptive Bisquare. Based on the analysis, this research revealed different model to every observations and different indicators. The eight locations are Paser, Kutai Kartanegara, West Kutai, East Kutai, Berau, Balikpapan, Samarinda dan Bontang. Those locations have variable that affected the morbidity number of dengue fever equally specifically house, elementary school facilities and public place that do not meet the requirements of health, and also waste transported  while for the observation location of Penajam Paser Utara has the affected variable of dengue fever morbidity number equally which are house, waste transported, elementary school facilities  and public place that do not meet the requirements of health, and also the citizen that do not have the healthy and hygienic lifestyle pattern.