arxivcs.LGstat.ML2026-06-30
Relational and Sequential Conformal Inference for Energy Time Series over Graphs via Foundation Models
Keivan Faghih Niresi, Alice Cicirello, Olga Fink
Accurate energy demand forecasting is essential for the reliable operation and planning of modern sustainable energy systems. Spatial-temporal graph neural networks (STGNNs) have recently achieved strong performance in point forecasting by jointly modeling temporal dynamics and r…