Physics-motivated dynamic spatio-temporal GNNs (EdgeGridNet) for industrial wind power forecasting (Baidu KDD Cup 2022). Research in dynamic graph learning, AI4Science, environmental forecasting, and resource-aware edge AI.
deep-learning multi-agent-systems spatio-temporal renewable-energy smart-grid time-series-forecasting graph-neural-networks edge-ai pytorch-geometric wind-power-forecasting physics-informed-ml ai4science dynamic-graph-learning environmental-forecasting resource-aware-ai hydrological-forecasting kdd-cup-2022
-
Updated
Sep 26, 2026 - Jupyter Notebook