Summary
This study applies advanced machine learning techniques (GBMs, RFs, SVMs, and ANNs) to predict maize biomass yield across United States agricultural regions, integrating engineered soil and environmental features. An ensemble approach combining GBMs, RFs, and ANNs achieved superior performance (RMSE 0.80 t/ha, R² 0.89), with SHAP analysis identifying Soil Fertility Index, Growing Degree Days, and cumulative rainfall as the most influential predictors. The work emphasises the utility of hybrid modelling for precision agriculture applications and rural environmental governance, whilst acknowledging the need for further investigation into cultural and political factors affecting farmer participation.
Regional applicability
The findings derive from United States data and may have limited direct applicability to United Kingdom cropping conditions, which differ substantially in climate, soil types, and growing season length. However, the methodological framework—feature engineering, ensemble modelling, and SHAP-based interpretability—could be adapted to UK maize production systems and other temperate cereals if locally calibrated with UK environmental and agronomic data.
Key measures
Biomass yield (t/ha); RMSE (0.80 t/ha); R² (0.89); Soil Fertility Index; Growing Degree Days; cumulative rainfall; SHAP feature importance values
Outcomes reported
The study developed and validated machine learning models to predict corn biomass yield using environmental, soil, and crop management data. Performance metrics included root mean squared error (RMSE) and coefficient of determination (R²) for ensemble and individual model approaches.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.