DeepCNNs-surrogate-UQ

The repository contains all files of the CACAIE paper entitled Deep convolutional neural networks for uncertainty propagation in random fields. Browse each of the folders for more information.

• The Examples folder provides the computer codes.
• The Responses folder provides peer reviews from 9 reviewers.
• The Manuscript folder provides the revised paper.

Summary

A machine learning approach is proposed for quantifying the effect of spatial variabilities in coupled elliptic systems. The learning model takes a hierarchical form where deep convolutional neural networks are used as the underlying components.

The learning process of this field-to-field mapping is efficient as distant connections among nonadjacent layers are established to improve the model efficiency in terms of training and deploying.

Citation

The paper has been accepted by the Journal of Computer-Aided Civil and Infrastructure Engineering. The below information contains the references of Journal.

@article{luo2019deep,
  title={Deep convolutional neural networks for uncertainty propagation in random fields},
  author={Luo, Xihaier and Kareem, Ahsan},
  journal={Computer-Aided Civil and Infrastructure Engineering},
  year={2019},
  publisher={Wiley Online Library}
}

Dependency

• Python 3.0
• TensorFlow 1.3
• Matplotlib 3.0
• Scipy 1.3
• Numpy 1.10

Dataset

We provide the data to help researchers reproduce the results. Please place the data in a proper directory.

Link to the datasets