/Forschungsprojekt

Numerical simulations are used in a variety of computational problems. The error of the simulation is often of interest here. However, the error is usually not known and must be calculated using expensive methods such as adjoint error equations. We investigate the use of multiple Convulational Neural Networks (CNNs) architectures as efficient and reliable error estimators for muscle fiber simulations by opendihu.

Primary LanguageJupyter NotebookMIT LicenseMIT

No issues in this repository yet.