TY - GEN
T1 - Global Surrogate Modeling by Neural Network-Based Model Uncertainty
AU - Leifsson, Leifur
AU - Nagawkar, Jethro
AU - Barnet, Laurel
AU - Bryden, Kenneth
AU - Koziel, Slawomir
AU - Pietrenko-Dabrowska, Anna
N1 - Funding Information: Acknowledgements. This material is based upon work supported in part by the Department of Energy under a Laboratory Directed Research and Development grant at Ames Laboratory. Publisher Copyright: © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - This work proposes a novel adaptive global surrogate modeling algorithm which uses two neural networks, one for prediction and the other for the model uncertainty. Specifically, the algorithm proceeds in cycles and adaptively enhances the neural network-based surrogate model by selecting the next sampling points guided by an auxiliary neural network approximation of the spatial error. The proposed algorithm is tested numerically on the one-dimensional Forrester function and the two-dimensional Branin function. The results demonstrate that global surrogate modeling using neural network-based function prediction can be guided efficiently and adaptively using a neural network approximation of the model uncertainty.
AB - This work proposes a novel adaptive global surrogate modeling algorithm which uses two neural networks, one for prediction and the other for the model uncertainty. Specifically, the algorithm proceeds in cycles and adaptively enhances the neural network-based surrogate model by selecting the next sampling points guided by an auxiliary neural network approximation of the spatial error. The proposed algorithm is tested numerically on the one-dimensional Forrester function and the two-dimensional Branin function. The results demonstrate that global surrogate modeling using neural network-based function prediction can be guided efficiently and adaptively using a neural network approximation of the model uncertainty.
KW - Error based exploration
KW - Global surrogate modeling
KW - Model uncertainty
KW - Neural networks
UR - https://www.scopus.com/pages/publications/85134290448
UR - https://doi.org/10.1007/978-3-031-08757-8_35
U2 - 10.1007/978-3-031-08757-8_35
DO - 10.1007/978-3-031-08757-8_35
M3 - Conference contribution
SN - 9783031087561
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 425
EP - 434
BT - Computational Science - ICCS 2022, 22nd International Conference, Proceedings
A2 - Groen, Derek
A2 - de Mulatier, Clélia
A2 - Krzhizhanovskaya, Valeria V.
A2 - Sloot, Peter M.A.
A2 - Paszynski, Maciej
A2 - Dongarra, Jack J.
PB - Springer Science and Business Media Deutschland GmbH
T2 - 22nd Annual International Conference on Computational Science, ICCS 2022
Y2 - 21 June 2022 through 23 June 2022
ER -