Predicting Maize Yield with Machine Learning
Author Information
Author(s): Ennaji Oumnia, Baha Sfia, Vergutz Leonardus, El Allali Achraf, Meher Prabina Kumar
Primary Institution: University Mohammed VI Polytechnic, Benguerir, Morocco
Hypothesis
Can gradient boosting methods improve yield estimation for the maize hybrid ZhengDan 958?
Conclusion
The study found that gradient boosting methods, particularly XGBoost, provide accurate yield predictions for the maize hybrid ZhengDan 958.
Supporting Evidence
- XGBoost achieved the highest prediction accuracy with an R2 of 0.87.
- Gradient Boosting Regressor also showed strong performance with an R2 of 0.86.
- Weather characteristics were found to significantly influence crop yield predictions.
Takeaway
This study shows that using smart computer programs can help farmers know how much corn they can grow based on weather and soil conditions.
Methodology
The study used various machine learning algorithms, including XGBoost and Gradient Boosting Regressor, to analyze over 1700 maize yield data pairs.
Potential Biases
Potential biases may arise from the dataset's regional focus and the specific environmental conditions of the study area.
Limitations
The study primarily focused on a single maize hybrid and may not generalize to other crops or hybrids.
Participant Demographics
The study focused on maize yield data from smallholder farms in the North China Plain.
Statistical Information
P-Value
10−16
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website