Deep neural networks (DNN) can model nonlinear relations between physical quantities.
Those DNNs are embedded in physical systems described by partial differential equations …

This book studies methods to concretely address inverse problems. An inverse problem
arises when the causes that produced a given effect must be determined or when one seeks …

Checkpointing techniques become more and necessary for the computation of adjoints. This
paper presents the more common multi-level checkpointing as well as the less known …

This paper proposes an approach based on automatic code differentiation to compute the
derivatives of a sizing model during the optimization of an electromagnetic device. After a …

Derivative-free optimization (DFO) has enjoyed renewed interest over the past years, mostly
motivated by the ever growing need to solve optimization problems defined by functions …

Les problèmes inverses sont omniprésents dans les sciences et l'ingénierie. Ils se
rencontrent à chaque fois que l'on cherche les causes ayant produit un effet connu ou que …

For derivative calculations, debugging, and interactive control one may need to reverse the
execution of a computer program for given inputs. If any increase of the time needed for the …

Full waveform inversion (FWI) is a promising technique that is capable of creating high-
resolution subsurface images of the earth from seismic data. However, it is computationally …

Velocity analysis resolves relatively long scales of earth structure, on the order of 1 km.
Migration produces images with length scales (wavelengths) on the order of 10's of m. In …

This thesis presents the implications of using adaptive time-step** schemes with the
adjoint-state method, a widely used algorithm for computing derivatives in optimal-control …