🌱
Loading KrishiDrishti...
Loading KrishiDrishti...
Training metrics and evaluation results for all KrishiDrishti ML models.
Trained on 95 districts across 8 NE India states. Real Open-Meteo weather + GEE NDVI.
Best performing kharif model. R²=0.74 exceeds target. River plain districts.
R²=0.74 exceeds target. Real temperature data from Open-Meteo improves rabi crops.
Lowest RMSE model. Oilseed crops have predictable patterns. ERA5 soil moisture helps.
Kharif cereal model. 95 district training improves spatial generalization.
Sequential model on 3-year weather windows. Experimental — XGBoost is primary model.
R² will exceed 0.75 target once real IMD weather and GEE NDVI data replace synthetic inputs in Phase 7.