Claim · model-written summary
Decision Trees had lower accuracy for climate-yield relationships.
Not reportedMaterialGeneCompoundOrgan / stageInteractorInteractor roleSecond interactorSecond interactor roleTreatmentSettingValueQualifier
Verbatim
e0“Neural Networks and Decision Trees showed less accuracy in capturing intricate relationships between climate variables and rice yields”
e1“The Random Forest model predicts Aman rice yields to reach 133.31 metric tonnes by 2030 (34.11 % of total rice production)”
Title
A comparative study of machine learning models for predicting Aman rice yields in Bangladesh
Taufiqul Islam; Tanmoy Mazumder; Md. Nishad Shahriair Roni; et al. Heliyon 2024. doi:10.1016/j.heliyon.2024.e40764
Other claims from this paper (10)
- Method / resource Random Forest was the most accurate rice-yield model.
- Method / resource Gradient Boosting performed well but had slightly higher error margins.
- Method / resource Linear Regression gave reasonable outputs but struggled with nonlinear patterns.
- Method / resource Neural Networks had lower accuracy for climate-yield relationships.
- Observation Random Forest projects 133.31 metric tonnes of Aman rice yield by 2030.
- Observation Random Forest projects 140 metric tonnes of Aman rice yield by 2050.
- Observation Projected temperatures rise by 2030 and 2050.
- Observation Projected monsoon rainfall increases slightly.
- Association The authors warn higher temperatures could stress rice production.
- Recommendation The authors recommend adaptive agricultural techniques and policies for rice production.