Bi-fidelity Variational Auto-encoder

This repo contains code for our paper

N. Cheng, O. A. Malik, S. Becker, A. Doostan. Bi-fidelity Variational Auto-encoder for Uncertainty Quantification. Computer Methods in Applied Mechanics and Engineering, 2024.

The paper is available at arXiv and DOI.

Referencing this code

If you use this code in any of your own work, please reference our paper:

@article{cheng2024bi,
  title={Bi-fidelity variational auto-encoder for uncertainty quantification},
  author={Cheng, Nuojin and Malik, Osman Asif and De, Subhayan and Becker, Stephen and Doostan, Alireza},
  journal={Computer Methods in Applied Mechanics and Engineering},
  volume={421},
  pages={116793},
  year={2024},
  publisher={Elsevier}
}

Description of code

Main file

• models.py: contain a bi-fidelity data class BFDataset, a fully-connected variational auto-encoder class, and a fully-connected bi-fidelity variational auto-encoder;

Helper file

• MMD_func.py: compute the maximum mean discrepancy (MMD) with a mixture of rational quadratic kernels;

Experiment files

• BF-VAE-beam.ipynb: BF-VAE implementation on the composite beam model with MMD results;

• BF-VAE-cav.ipynb: BF-VAE implementation on the cavity flow model with MMD results;

• BF-VAE-burgers.ipynb: BF-VAE implementation on the 1D visous Burgers model with MMD results.

Data

The bi-fidelity data is available in this link.

Author contact information

Please feel free to contact me at any time if you have any questions or would like to provide feedback on this code or on the paper. I can be reached at nuojin (dot) cheng (at) colorado (dot) edu.