Bharadwaj, P and Li, M and Demanet, L (2022) Redatuming physical systems using symmetric autoencoders. In: Physical Review Research, 4 (2).
|
PDF
phy_rev_res_4-2_2022 .pdf - Published Version Download (2MB) | Preview |
Abstract
This paper considers physical systems described by hidden states and indirectly observed through repeated measurements corrupted by unmodeled nuisance parameters. A network-based representation learns to disentangle the coherent information (relative to the state) from the incoherent nuisance information (relative to the sensing). Instead of physical models, the representation uses symmetry and stochastic regularization to inform an autoencoder architecture called SymAE. It enables redatuming, i.e., creating virtual data instances where the nuisances are uniformized across measurements. © 2022 authors. Published by the American Physical Society.
| Item Type: | Journal Article |
|---|---|
| Publication: | Physical Review Research |
| Publisher: | American Physical Society |
| Additional Information: | The copyright for this article belongs to the American Physical Society. |
| Keywords: | Stochastic models; Stochastic systems, Auto encoders; Hidden state; Learn+; Network-based; Nuisance parameter; Physical modelling; Physical systems; Repeated measurements; Stochastics; Symmetrics, Learning systems |
| Department/Centre: | Division of Mechanical Sciences > Centre for Earth Sciences |
| Date Deposited: | 29 Jun 2022 07:33 |
| Last Modified: | 29 Jun 2022 07:33 |
| URI: | https://eprints.iisc.ac.in/id/eprint/73968 |
Actions (login required)
 |
View Item |
