Current browse context:
physics.comp-ph
Change to browse by:
References & Citations
Physics > Computational Physics
Title: Learning and correcting non-Gaussian model errors
(Submitted on 29 May 2020 (v1), last revised 27 Jan 2021 (this version, v2))
Abstract: All discretized numerical models contain modelling errors - this reality is amplified when reduced-order models are used. The ability to accurately approximate modelling errors informs statistics on model confidence and improves quantitative results from frameworks using numerical models in prediction, tomography, and signal processing. Further to this, the compensation of highly nonlinear and non-Gaussian modelling errors, arising in many ill-conditioned systems aiming to capture complex physics, is a historically difficult task. In this work, we address this challenge by proposing a neural network approach capable of accurately approximating and compensating for such modelling errors in augmented direct and inverse problems. The viability of the approach is demonstrated using simulated and experimental data arising from differing physical direct and inverse problems.
Submission history
From: Danny Smyl [view email][v1] Fri, 29 May 2020 14:20:17 GMT (17658kb,D)
[v2] Wed, 27 Jan 2021 08:12:07 GMT (19957kb,D)
Link back to: arXiv, form interface, contact.