Comparison of spatial adjacency matrices (Rook, Queen) of SAR-MA model using MLE for diabetes

This study aims to analyze the effect of spatial dependence on the estimates of the autoregressive-spatial moving averages (SAR-MA) model using the maximum likelihood method (MLE) and to test the effect of different adjacency matrices (Rook and Queen) on the accuracy of the model. The study was based on real data related to diabetes. among children in Iraq for the year 2024, where data was collected from various governorates, including variables such as age, gender, cumulative blood sugar level, and other factors that affect the spread of the disease, Moran's Z test was used to test the presence of spatial dependence, followed by the process of estimating the model parameters using the maximum likelihood method (MLE). The results showed a noticeable spatial dependence in some variables (S4, S5, S8) while it did not appear in others, It was also shown that the use of the Rook adjacency matrix significantly improved the accuracy of the model, so the study recommended using the SAR-MA model with the Rook adjacency matrix in spatial analysis to ensure accurate estimates, and to expand the scope of future applications of spatial models in areas such as economics and urban planning.