Predictive modeling of regional economic growth using agricultural and socioeconomic indicators

Abstract

Economic growth and food security are closely interconnected dimensions of sustainable regional development, particularly in agrarian regions such as Lampung Province, Indonesia. However, conventional analytical approaches often fail to capture the complex and nonlinear relationships between agricultural productivity and socio-economic conditions. This study aims to analyze the determinants of regional economic growth by integrating agricultural and socio-economic indicators using a random forest–based modeling framework. Secondary panel data from 15 districts over the period 2014–2024 were analyzed, comprising 165 observations and 14 explanatory variables. The results show that agricultural production plays a dominant role, with rice production and harvested area contributing approximately 39.8% and 34.7% of total feature importance, respectively. The model achieved strong predictive performance with a coefficient of determination (R2) of 0.68 and root mean squared error (RMSE) of 6,346.82, indicating that the selected variables explain a substantial portion of gross regional domestic product (GRDP) variation. Socio-economic factors, including poverty rate, per capita expenditure, and human development index (HDI), also contribute meaningfully to regional economic outcomes. These findings highlight the importance of integrating agricultural productivity with social development policies to achieve inclusive and sustainable economic growth. The proposed approach provides a data-driven framework to support regional policy formulation and improve food security strategies.