A training strategy for hybrid models to break the curse of dimensionality

Mechanistic/data-driven hybrid modeling is a key approach when the mechanistic details of the processes at hand are not sufficiently well understood, but also inferring a model purely from data is too complex. By the integration of first principles into a data-driven approach, hybrid modeling promis...

Descripción completa

Detalles Bibliográficos
Autores principales: E. Samadi, Moein, Kiefer, Sandra, Fritsch, Sebastian Johaness, Bickenbach, Johannes, Schuppert, Andreas
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Public Library of Science 2022
Materias:
Research Article
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9477345/
https://www.ncbi.nlm.nih.gov/pubmed/36107916
http://dx.doi.org/10.1371/journal.pone.0274569

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9477345/
https://www.ncbi.nlm.nih.gov/pubmed/36107916
http://dx.doi.org/10.1371/journal.pone.0274569

Ejemplares similares