In the framework of reduced basis methods, we recently introduced a new certified hierarchical and adaptive surrogate model, which can be used for efficient approximation of input-output maps that are governed by parametrized partial differential equations. This adaptive approach combines a full order model, a reduced order model and a machine-learning model. In this contribution, we extend the approach by leveraging novel kernel models for the machine learning part, especially structured deep kernel networks as well as two layered kernel models. We demonstrate the usability of those enhanced kernel models for the RB-ML-ROM surrogate modeling chain and highlight their benefits in numerical experiments.
%0 Conference Paper
%1 10.1007/978-3-031-56208-2_11
%A Wenzel, Tizian
%A Haasdonk, Bernard
%A Kleikamp, Hendrik
%A Ohlberger, Mario
%A Schindler, Felix
%B Large-Scale Scientific Computations
%C Cham
%D 2024
%E Lirkov, Ivan
%E Margenov, Svetozar
%I Springer Nature Switzerland
%K unibibliografie haasdonk wenzel PN6 EXC2075 ians anm fis
%P 117--125
%T Application of Deep Kernel Models for Certified and Adaptive RB-ML-ROM Surrogate Modeling
%X In the framework of reduced basis methods, we recently introduced a new certified hierarchical and adaptive surrogate model, which can be used for efficient approximation of input-output maps that are governed by parametrized partial differential equations. This adaptive approach combines a full order model, a reduced order model and a machine-learning model. In this contribution, we extend the approach by leveraging novel kernel models for the machine learning part, especially structured deep kernel networks as well as two layered kernel models. We demonstrate the usability of those enhanced kernel models for the RB-ML-ROM surrogate modeling chain and highlight their benefits in numerical experiments.
%@ 978-3-031-56208-2
@inproceedings{10.1007/978-3-031-56208-2_11,
abstract = {In the framework of reduced basis methods, we recently introduced a new certified hierarchical and adaptive surrogate model, which can be used for efficient approximation of input-output maps that are governed by parametrized partial differential equations. This adaptive approach combines a full order model, a reduced order model and a machine-learning model. In this contribution, we extend the approach by leveraging novel kernel models for the machine learning part, especially structured deep kernel networks as well as two layered kernel models. We demonstrate the usability of those enhanced kernel models for the RB-ML-ROM surrogate modeling chain and highlight their benefits in numerical experiments.},
added-at = {2024-08-02T11:24:46.000+0200},
address = {Cham},
author = {Wenzel, Tizian and Haasdonk, Bernard and Kleikamp, Hendrik and Ohlberger, Mario and Schindler, Felix},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2def1b46e124d6b63c807d1a4cb6815a1/mathematik},
booktitle = {Large-Scale Scientific Computations},
editor = {Lirkov, Ivan and Margenov, Svetozar},
interhash = {862f1092efd9081fd598d313fcfb8a40},
intrahash = {def1b46e124d6b63c807d1a4cb6815a1},
isbn = {978-3-031-56208-2},
keywords = {unibibliografie haasdonk wenzel PN6 EXC2075 ians anm fis},
pages = {117--125},
publisher = {Springer Nature Switzerland},
timestamp = {2024-08-02T11:24:46.000+0200},
title = {Application of Deep Kernel Models for Certified and Adaptive RB-ML-ROM Surrogate Modeling},
year = 2024
}
