International agricultural trade forecasting using machine learning

Abstract Focusing on seven major agricultural commodities with a long history of trade, this study employs data-driven analytics to decipher patterns namely using supervised machine learning (ML), as well neural networks. The ML and network techniques are trained data until 2010 2014, respectively. Results show the high relevance models forecasting trade in near- long-term relative traditional approaches, which often subjective assessments or time-series projections. While quantified key economic factors underlying flows, approaches provide better fits over term.