Answers are grounded in this paper and cite their sources.
Accurately predicting crop yields is fundamental to ensuring food security and formulating sustainable agricultural policies. In countries like Bangladesh, where environmental, agricultural, and regional variables are deeply interdependent, developing consistent and reliable yield models remains a major challenge due to limited data availability. Addressing this issue, the study introduces a novel machine learning framework that fuses numerical data (e.g., meteorological parameters, soil fertility indices, and production metrics) with textual data (e.g., soil reports and farming practice records) to forecast rice yield in the Chittagong district. Multiple regression and ensemble learning techniques were designed, validated, and compared to identify the most reliable predictor. Results indicate that the optimized weighted ensemble approach outperformed all others, attaining an R² value of 0.991, a mean absolute error of 0.026, and a root mean square error of 0.051. These outcomes highlight the effectiveness of integrating structured and unstructured features for yield prediction. The proposed methodology offers a scalable foundation for extending predictive modeling to diverse crops and regions, supporting data-informed and adaptive agricultural management.
No takes yet. Share an insight, caveat, or question.
Jabed et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: