Title: Modeling The Food Security Index In Indonesia Using The Mixed Geographically Weighted Regression (MGWR) Method
Authors: Aurellia Calista Anggakusuma, Toha Saifudin, Nur Chamidah, Elly Ana
Volume: 9
Issue: 12
Pages: 273-279
Publication Date: 2025/12/28
Abstract:
Food security is a strategic development issue characterized by differences in conditions between regions in Indonesia. These differences are reflected in the Food Security Index (FSI), which shows spatial variations between provinces. This study aims to model the Food Security Index in Indonesia in 2024 and identify the economic, social, and demographic factors that influence it, taking into account spatial heterogeneity. The method used is Mixed Geographically Weighted Regression (MGWR), which is able to distinguish between the influence of global and local variables. The data used is secondary data from the Central Statistics Agency and the National Food Agency with 38 provinces in Indonesia as the unit of analysis. The explanatory variables used include Gross Regional Domestic Product (GRDP) per capita, Human Development Index (HDI), Open Unemployment Rate (OUR), population, and number of poor people. The results of the study show autocorrelation and spatial heterogeneity, making a global regression model less suitable. Based on a comparison of AIC values and determination coefficients, the MGWR model provides the best performance with the lowest AIC value of 225.379 and a determination coefficient of 85.8%. The variables of GRDP per capita and population are global in nature, while HDI, OUR, and the numbers of poor people are local in nature with varying influences between provinces. The MAPE value of 4.27% indicates that the MGWR model has excellent prediction accuracy. The results of this study are expected to form the basis for the formulation of food security policies that are more targeted and based on regional characteristics, in line with efforts to achieve Sustainable Development Goal (SDG) 2: Zero Hunger.