IMPLEMENTASI METODE GEOGRAPHICALLY WEIGHTED REGRESSION (GWR) PADA KASUS DIARE BALITA DI PROVINSI JAWA TIMUR

Abstract

Spatial regression is an extension of classical regression analysis by considering spatial elements of spatial elements. One of the model of spatial regressions is the Geographically Weighted Regression (GWR). In the analysis, the GWR method considers the differences in characteristics between regions (spatial heterogeneity). Diarrhea cases in toddlers can be modeled using the GWR model. This research aims to model and identify factors that significantly influence diarrhea cases in toddlers in each district in East Java Province in 2020 using GWR. There are two weighting functions used in this research that are fixed bisquare kernel and adaptive bisquare kernel. The results showed that the GWR model with the adaptive kernel bisquare weighting function was more suitable because it produced the highest 𝑅 2 value of 79.29%. The factors that have a significant effect in each district are different and the dominant factor is the provision of vitamin A to toddlers.