Combining upwind difference matrices with Saint-Venant and Aw-Rascle discretizations in graph message passing improves flux prediction and edge-direction sensitivity on two real-world networks.
Stable solution of inverse problems
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Topology-aware Neural Flux Prediction Guided by Physics
Combining upwind difference matrices with Saint-Venant and Aw-Rascle discretizations in graph message passing improves flux prediction and edge-direction sensitivity on two real-world networks.