Researchers built a machine learning model that predicts rice yields in Tamil Nadu, India, by isolating the effect of fertilizer from weather, irrigation, and location. It predicted yields accurately and showed nitrogen matters most, with a few districts responding especially strongly to fertilizer.
1
The Fertilizer-Focused Stacked Ensemble Model reached R2 of 0.9512, RMSE of 0.0406, MAE of 0.0287, and MAPE of 0.202 on Tamil Nadu district-level paddy yield data.
2
It outperformed existing ensemble models by combining LightGBM, XGBoost, and a multilayer perceptron with a Gradient Boosting meta-learner after removing climate, irrigation, and spatial effects.
3
SHAP analysis ranked nitrogen as the most important fertilizer component and flagged Theni, Dharmapuri, Thiruvarur, and Dindigul as the most fertilizer-sensitive districts.
mail
Weekly plant science — one email, Saturdays.
Check your inbox to confirm — link expires in 24 hours.
Something went wrong — please try again.
Too many signup attempts from your network. Try again in an hour, or email hello@plant.news.
