chaztikov / caae-drdcn-inverse Goto Github PK

View Code? Open in Web Editor NEW

Deep residual networks for dimensionality reduction and surrogate modeling in high-dimensional inverse problems

Python 92.94% MATLAB 7.06%

caae-drdcn-inverse's Introduction

Integration of adversarial autoencoders with residual dense convolutional networks for estimation of non-Gaussian conductivities

Update (2021-3-13): We have now uploaded the codes of 3D-DRDCN. For 2D problems, one can replace the 3D opterations (e.g., convolution and batchnorm) with their 2D counterparts in net design file (i.e. rrde_ed3D.py).

Integration of adversarial autoencoders with residual dense convolutional networks for estimation of non-Gaussian conductivities

Shaoxing Mo, Nicholas Zabaras, Xiaoqing Shi, Jichun Wu

Dependencies

python 3
PyTorch 0.4
h5py
matplotlib
seaborn

Citation

See Mo et al. (2020) for more information. If you find this repo useful for your research, please consider to cite:

@article{doi:10.1029/2019WR026082,
author = {Mo, Shaoxing and Zabaras, Nicholas and Shi, Xiaoqing and Wu, Jichun},
title = {Integration of adversarial autoencoders with residual dense convolutional networks for estimation of 
non-Gaussian hydraulic conductivities},
journal = {Water Resources Research},
volume = {56},
number = {},
pages = {e2019WR026082},
doi = {10.1029/2019WR026082},
url = {https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2019WR026082}
}

or:

Mo, S., Zabaras, N., Shi, X., & Wu, J. ( 2020). Integration of adversarial autoencoders with residual dense 
convolutional networks for estimation of non‐Gaussian hydraulic conductivities. Water Resources Research, 
56, e2019WR026082. https://doi.org/10.1029/2019WR026082