Wheat Yield Prediction Based on Random Forest Method

Crop yield prediction is crucial for enhancing productivity, managing risks, ensuring food security, and improving the overall sustainability of agriculture. This study evaluated the Random Forest (RF) machine learning method for its ability to predict wheat yield responses to climate and soil related factors, with Multiple Linear Regression (MLR) serving as a benchmark. The results revealed that RF outperformed MLR in predicting wheat yield. Mean absolute error (MAE) and root mean squared error (RMSE) were used as evaluation metrics, with RF achieving an MAE of 135.88 and an RMSE of 163.90. In comparison, MLR produced an MAE of 435.74 and an RMSE of 653.39. These findings demonstrate that RF has superior wheat yield forecasting capabilities compared to MLR.