Geospatial-AI corn yield forecasting for the U.S. Corn Belt. Team project fine-tuning NASA/IBM Prithvi-EO-2.0-600M with LoRA, fused with weather/soil/drought features and calibrated uncertainty cones. State-level RMSE ~3–5 bu/ac — competitive with USDA WASDE. CSU Geospatial AI Hackathon 2026.
agriculture hackathon pytorch remote-sensing lora precision-agriculture geospacial crop-yield-prediction streamlit prithvi foundation-models nasa-hls
-
Updated
Apr 30, 2026 - Jupyter Notebook