Enhanced Crop Yield Predictions Amid Climate Change to Improve Agriculture With SwinTrans-Att Based Model

Harvest prediction is crucial to global, national, and local farming decisions. Soil, climate, environmental conditions, and crop characteristics affect harvest forecasts. Crop sensing, monitoring, management information systems, and variable rate technologies enable precision agriculture to enhance crop output and quality while minimizing environmental impact. The suggested method involves preprocessing, feature selection, and model training. Preprocessing normalizes input variables using min-max scaling to preserve distributions. Feature selection considers irrigation, land use, fertilizer application, and weather, which affect agricultural productivity. To trained the model with SwinTrans-Att, which handles complex agricultural data. The SwinTrans-Att model outperformed CNN and SVM with an average accuracy of 93.61%. This highlights its power for precision farming and crop management decisions. This study found that preprocessing, feature selection, and advanced modeling increase crop yield estimates. Precision agriculture's revolutionary potential to boost production and sustainability is shown