Volume 2 - Issue 3, May - June 2026

Agriculture contributes approximately 14% of global greenhouse gas emissions and is simultaneously the sector most vulnerable to climate variability and change. As extreme weather events intensify and growing seasons shift, agribusiness production systems require robust, forwardlooking analytical tools capable of transforming vast multisource environmental datasets into actionable operational intelligence. Machine learning offers precisely such a capability, enabling predictive modelling of crop yields, climate hazard risks, soil conditions, and supply chain vulnerabilities at unprecedented spatial and temporal resolutions. This paper presents a comprehensive review of machine learning driven predictive analytics frameworks and their applications across the dimensions of climate-smart agribusiness production systems, encompassing crop yield forecasting, climate hazard early warning, soil and water resource management, greenhouse gas emissions prediction, and supply chain optimisation. A systematic narrative review was conducted drawing exclusively on peer-reviewed academic literature from Scopus, Web of Science, PubMed, Nature, Frontiers, Springer, ScienceDirect, and the Consensus academic search repository. Thirty high-quality references spanning agricultural informatics, climate science, remote sensing, and agribusiness were synthesised. Evidence demonstrates that Long Short-Term Memory networks with attention mechanisms explain up to 73% of the spatiotemporal variance in crop yield under climate variability. Random Forest and Gradient Boosting models achieve prediction accuracies of 95.5% to 95.8% in IoT-enabled precision agriculture deployments. Machine learning integration into smart farming platforms reduces water consumption by 15%, boosts yield by 20%, and generates approximately USD 5,000 in annual pesticide savings per farm. Expert-driven explainable AI systems reliably detect multi-hazard agricultural climate risks, enhancing proactive adaptation. Barriers to adoption, particularly for smallholder farmers in developing economies, remain significant. Machine learning driven predictive analytics represents a transformative technological force for climate-smart agribusiness. Its full potential is contingent on addressing data infrastructure deficits, ensuring model explainability, and designing inclusive policy frameworks that prevent smallholder exclusion from the precision agriculture revolution.

Oluwasegun Alex Adepetoye, Jatto Suleman,"Machine Learning Driven Predictive Analytics for Climate-Smart Agribusiness Production Systems" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(3): Page(997-1004) May-June 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.