L. N. Reddy

Enhancing crop recommendation through graph attention and tabular deep learning on soil–climate datasets

The findings demonstrate that GeoTab-CRS provides a practical, deployable decision support tool for agricultural planners and state advisory systems, achieving real-time inference at 3.2 ms/sample with uncertainty quantification that enables risk-aware crop recommendations.