Data-Driven Predication of Wheat Production across Africa, the Middle East, and Oceania using Machine Learning Techniques

Abstract

Wheat is a crop that is consumed and produced globally. It is used in many food items in our daily life. Therefore, its widespread use globally and its share in food grains makes its extremely important. Therefore, considering the importance of the wheat crop, it is very important to estimate its production level in advance so that equitable availability of this important grain can be ensured at the global level. This study addresses the limitations of traditional statistical models by incorporating advanced machine learning techniques capable of capturing nonlinear relationships in agricultural data. Experimental results demonstrate that XGBoost outperforms other models with the accuracy level 99.99%. and an RMSE of 18.37 and MAE of 13.96, indicating high predictive accuracy. The findings highlight the potential of machine learning-driven approaches for supporting data-informed agricultural planning and policy formulation in regions with high production variability. This framework will not only help predict wheat production in advance but also serve as a useful tool for agricultural policymaking.