Harvesting Insights: Unveiling the Future of Mango Yield Using Real-Time Bangladesh Climate Data by Supervised Machine Learning Approach

Mango, a unique fruit of the Indian subcontinent, was cultivated under the supervision of Mughal emperor Akbar [1]. Its exceptional taste impressed Alexander the Great [2]. Considering the potentiality of this unique fruit this study focuses on predicting mango yield in Bangladesh using machine learning algorithms. Bangladesh's economy heavily relies on the export of Ready-made Garments (RMG) and Knitwear, and there is a need to explore alternative revenue sources due to the increasing trade deficit. Mango production has the potential to contribute significantly to export earnings and GDP growth. However, predicting mango yield has become challenging due to climate change and natural calamities. To address this, a comprehensive dataset covering 54 years of mango production and weather conditions in all 64 districts of Bangladesh was collected. Regression algorithms were employed to predict mango yield, with the Random Forest Regression (RFR) model exhibiting the lowest error rates. The contributions and practical impact of this study include the creation of a real-time dataset, demonstrating the accuracy of the proposed algorithms, and facilitating improved health outcomes and economic benefits through effective production planning.