ML-driven agent-based simulation of agri-food supply chain

Abstract

The paper presents the joint use of machine learning methods and agent-based modeling for analysis and scenario forecasts to reduce the impact of trade flow destabilization on food security in Russia in increasing sanctions pressure. The authors propose a conceptual scheme of a software and analytical complex for forecasting indicators of agri-food supply chains. The obtained results can form the basis of a socio-economic multi-agent model for ensuring food security. Using the proposed approach in the situation centers can help counter external threats and ensure Russia's national security.