PI-DIONs train DeepONet-style networks to invert PDEs using only physics residuals and partial measurements, with no labeled (measurement, parameter) pairs.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Physics-Informed Deep Inverse Operator Networks for Solving PDE Inverse Problems
PI-DIONs train DeepONet-style networks to invert PDEs using only physics residuals and partial measurements, with no labeled (measurement, parameter) pairs.