A unified framework with FPI-BPINN and fParVI-PINN approaches enables functional priors in Bayesian PINN inversion, yielding accurate posterior estimates for 1D seismic tomography and 2D Darcy flow permeability inversion.
Journal of Computational Physics 425, 109913
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.geo-ph 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Functional-prior-based approaches to Bayesian PDE-constrained inversion using physics-informed neural networks
A unified framework with FPI-BPINN and fParVI-PINN approaches enables functional priors in Bayesian PINN inversion, yielding accurate posterior estimates for 1D seismic tomography and 2D Darcy flow permeability inversion.