Robust spatial Durbin approach in modelling the spreading of tuberculosis in Indonesia

Abstract

Spatial regression analysis is a method used for data that has spatial effects. One method in spatial regression analysis is the Spatial Durbin Model (SDM) which shows spatial effects on both dependent and independent variables. In the spatial regression model, inaccuracy can occur in predicting the model due to spatial outliers. To overcome outliers in the SDM model, a robust method is needed, namely the Robust Spatial Durbin Model (RSDM). This study was conducted to model the factors that influence the spread of tuberculosis (TB) cases in Indonesia and determine the best method. The results obtained are that the RSDM model is better with a larger Adjusted R² value and a smaller Mean Squared Error (MSE) value than SDM. The factors suspected of influencing the spread of TB cases in Indonesia are the percentage of households with access to proper sanitation, the number of HIV cases and population density.