BEC-NN

Neural network for Born effective charge prediction.

This repository contains:

• Source code for training a neural network that predicts Born effective charges (BEC)
• Models used for the publication: Study of the ion mobility in defect-laden ZrO₂ under an electric field using neural network with predictions for Born effective charges (DOI: 10.1103/jcsd-dbl2)
• Lammps interface for running simulations with BEC-NN
• Data set (10,103 structures)

Citation

If you use the code, models, or dataset in this repository, please cite the corresponding paper:

Study of the ion mobility in defect-laden ZrO₂ under an electric field using neural network with predictions for Born effective charges Anh Khoa Augustin Lu, Naoki Maekawa, Akane Ikeda, Koji Shimizu, Hiroshi Masuda, Hidehiro Yoshida, and Satoshi Watanabe.
Physical Review Materials (Accepted May 4, 2026).
DOI: 10.1103/jcsd-dbl2

BibTeX

@article{Lu_2026,
  title = {Study of the ion mobility in defect-laden {ZrO$_2$} under an electric field using neural network with predictions for Born effective charges},
  author = {Lu, Anh Khoa Augustin and Maekawa, Naoki and Ikeda, Akane and Shimizu, Koji and Masuda, Hiroshi and Yoshida, Hidehiro and Watanabe, Satoshi},
  journal = {Physical Review Materials},
  year = {2026},
  doi = {10.1103/jcsd-dbl2},
  url = {https://doi.org/10.1103/jcsd-dbl2},
  note = {Accepted May 4, 2026}
}

License

This project is licensed under the MIT License — see the LICENSE file for details.

SPDX identifier: MIT Copyright (c) 2025-2026 Anh Khoa Augustin Lu

Maintainer

Anh Khoa Augustin Lu — lu.augustin@nims.go.jp

GitHub: https://github.com/AugustinLu

Issue tracker: https://github.com/AugustinLu/BEC-NN/issues