Maya Adelia Sari, Trianingsih Eni Lestari
Welfare and development in a region are crucial because they are interrelated in determining the quality of life of its people. The Human Development Index (HDI) and the poverty rate are indicators that can be used to identify this. The SUR-SEM model is a combination of two different regression models, namely the Seemingly Unrelated Regression (SUR) and the Spatial Error Model (SEM). This study focuses on HDI data and poverty rates from 38 districts/cities in East Java in 2022. In this study, parameter estimation for the SUR-SEM model was performed using MLE estimation. Additionally, this study also uses Queen Contiguity spatial weighting and Customize spatial weighting. The parameter estimation conducted still results in non-close form equations for the parameters λ and σ, necessitating further approaches using Newton-Raphson iteration. The results of the SUR-SEM modeling in this study show that the use of Queen Contiguity spatial weighting produces a model with a higher R-square value and the smallest RMSE. The HDI obtained an R-square value of 97.37% and an R-square value of 75.93% for the poverty rate. © 2025 American Institute of Physics Inc.. All rights reserved.
Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Negeri Malang, Malang, Indonesia
