Spatial Regression in the Number of Covid-19 Cases in Bandung City

Abstract

Regression method that pays attention to neighborhood or region is Spatial Regression with Spatial Autoregressive (SAR) and Spatial Error Model (SEM) approaches. This method is used as an approach for cases of the spread of Covid-19 , namely the number of active Covid-19 patients in the city of Bandung with the independent variable in the form of population per hectare. , population age 60 and over, poverty and population not working. The results of the Moran index test show that there is a spatial effect on the spread of positive active Covid-19 patients, and the SAR model is better than the classical linear regression model (OLS; Ordinary Least Square) and SEM with variables/factors having a significant effect, namely population 60 and over, poverty and residents do not work. The AIC value of the SEM model is 262.845 with a coefficient of determination (R-squared) of 0.457302. Keywords: Spatial Regression, OLS, SAR, AIC, Coefficient of Determination.