MLP-Mixer based surrogate model for seismic ground motion with spatial source and geometry parameters

Hirotaka Hachiya, Yuto Kuroki, Asako Iwaki, Takahiro Maeda, Naonori Ueda, Hiroyuki Fujiwara
Proceedings of the 16th Asian Conference on Machine Learning, PMLR 260:191-206, 2025.

Abstract

Seismic motion simulations enable high-precision predictions but are computationally demanding. This study introduces a deep learning surrogate model using the MLP-Mixer architecture to address this challenge. Traditional models using independent Multi-layer Perceptrons (MLPs) fail to capture spatial correlations, while U-shaped Neural Operators (U-NOs) require high computational costs for high-resolution inputs and outputs. Our proposed model, the Multiple MLP-Mixer (Multi-MLP-Mixer), integrates global and local spatial information through multiple MLP-Mixer blocks and dual patch-wise affine transformations. We demonstrate the effectiveness of our method with simulation data from anticipated megathrust earthquakes in the Nankai Trough, achieving performance comparable to state-of-the-art models with significantly improved computational efficiency.

Cite this Paper

BibTeX
@InProceedings{pmlr-v260-hachiya25a, title = {MLP-Mixer based surrogate model for seismic ground motion with spatial source and geometry parameters}, author = {Hachiya, Hirotaka and Kuroki, Yuto and Iwaki, Asako and Maeda, Takahiro and Ueda, Naonori and Fujiwara, Hiroyuki}, booktitle = {Proceedings of the 16th Asian Conference on Machine Learning}, pages = {191--206}, year = {2025}, editor = {Nguyen, Vu and Lin, Hsuan-Tien}, volume = {260}, series = {Proceedings of Machine Learning Research}, month = {05--08 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v260/main/assets/hachiya25a/hachiya25a.pdf}, url = {https://proceedings.mlr.press/v260/hachiya25a.html}, abstract = {Seismic motion simulations enable high-precision predictions but are computationally demanding. This study introduces a deep learning surrogate model using the MLP-Mixer architecture to address this challenge. Traditional models using independent Multi-layer Perceptrons (MLPs) fail to capture spatial correlations, while U-shaped Neural Operators (U-NOs) require high computational costs for high-resolution inputs and outputs. Our proposed model, the Multiple MLP-Mixer (Multi-MLP-Mixer), integrates global and local spatial information through multiple MLP-Mixer blocks and dual patch-wise affine transformations. We demonstrate the effectiveness of our method with simulation data from anticipated megathrust earthquakes in the Nankai Trough, achieving performance comparable to state-of-the-art models with significantly improved computational efficiency.} }
Endnote
%0 Conference Paper %T MLP-Mixer based surrogate model for seismic ground motion with spatial source and geometry parameters %A Hirotaka Hachiya %A Yuto Kuroki %A Asako Iwaki %A Takahiro Maeda %A Naonori Ueda %A Hiroyuki Fujiwara %B Proceedings of the 16th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Vu Nguyen %E Hsuan-Tien Lin %F pmlr-v260-hachiya25a %I PMLR %P 191--206 %U https://proceedings.mlr.press/v260/hachiya25a.html %V 260 %X Seismic motion simulations enable high-precision predictions but are computationally demanding. This study introduces a deep learning surrogate model using the MLP-Mixer architecture to address this challenge. Traditional models using independent Multi-layer Perceptrons (MLPs) fail to capture spatial correlations, while U-shaped Neural Operators (U-NOs) require high computational costs for high-resolution inputs and outputs. Our proposed model, the Multiple MLP-Mixer (Multi-MLP-Mixer), integrates global and local spatial information through multiple MLP-Mixer blocks and dual patch-wise affine transformations. We demonstrate the effectiveness of our method with simulation data from anticipated megathrust earthquakes in the Nankai Trough, achieving performance comparable to state-of-the-art models with significantly improved computational efficiency.
APA
Hachiya, H., Kuroki, Y., Iwaki, A., Maeda, T., Ueda, N. & Fujiwara, H.. (2025). MLP-Mixer based surrogate model for seismic ground motion with spatial source and geometry parameters. Proceedings of the 16th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 260:191-206 Available from https://proceedings.mlr.press/v260/hachiya25a.html.

Related Material