Dani Shofi Nurizza
UPN 'Veteran' East Java, Surabaya

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Geo AI-Based Food Policy Navigation in South Sumatra Dian Rizky; Badara Shofi Dana; Dani Shofi Nurizza
Indonesian Journal of Islamic Economics and Finance Vol. 6 No. 2 (2026)
Publisher : Institut Agama Islam Sunan Giri Ponorogo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37680/ijief.v6i2.10504

Abstract

Food price instability in South Sumatra is structural and recurring, exceeding the national inflation rate. This study develops a Geo AI-based food policy navigation system integrating satellite biophysical indicators, macroeconomic data, and machine learning to support the Regional Inflation Control Team (TPID). The data comprise a monthly panel from 11 of 17 districts (March 2021 to December 2025), selected based on available retail price data. Methods include GSTAR, SAR-X, SARIMAX, and XGBoost for forecasting; SHAP, Granger causality, and VAR decomposition for root cause diagnostics; PCA for the Composite Food Security Index (IKP Sumsel); Monte Carlo simulation for shock scenarios; and MCDA-SAW for policy prioritization. GSTAR achieves the highest accuracy (MAPE 2.56% for rice, 3.36% for eggs, 4.34% for chicken). Root causes vary across commodities, with 54.55% of districts under Alert status by December 2025 and karhutla as the most significant threat. The system integrates real-time satellite signals, including NDVI, rainfall, and forest fire hotspots, directly into TPID decision-making to enable anticipatory rather than reactive interventions. The study recommends differentiated 4K Strategy priorities for Musi Banyuasin, Ogan Komering Ilir, and Musi Rawas, and fire mitigation integrated into peatland food security policy.