This record has versions v1, v2, v3. This is version v2.
×

Recommended by

Indexed by

Adaptive energy reference for machine-learning models of the electronic density of states

Wei Bin How1*, Sanggyu Chong1*, Federico Grasselli1*, Kevin K. Huguenin-Dumittan1*, Michele Ceriotti1*

1 Laboratory of Computational Science and Modeling (COSMO), IMX, École Polytechnique Fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland

* Corresponding authors emails: weibin.how@epfl.ch, sanggyu.chong@epfl.ch, federico.grasselli@epfl.ch, kevin.huguenin-dumittan@epfl.ch, michele.ceriotti@epfl.ch
DOI10.24435/materialscloud:y6-m4 [version v2]

Publication date: Oct 11, 2024

How to cite this record

Wei Bin How, Sanggyu Chong, Federico Grasselli, Kevin K. Huguenin-Dumittan, Michele Ceriotti, Adaptive energy reference for machine-learning models of the electronic density of states, Materials Cloud Archive 2024.158 (2024), https://doi.org/10.24435/materialscloud:y6-m4

Description

The electronic density of states (DOS) provides information regarding the distribution of electronic states in a material, and can be used to approximate its optical and electronic properties and therefore guide computational material design. Given its usefulness and relative simplicity, it has been one of the first electronic properties used as target for machine-learning approaches going beyond interatomic potentials. A subtle but important point, well-appreciated in the condensed matter community but usually overlooked in the construction of data-driven models, is that for bulk configurations the absolute energy reference of single-particle energy levels is ill-defined. Only energy differences matter, and quantities derived from the DOS are typically independent on the absolute alignment. We introduce an adaptive scheme that optimizes the energy reference of each structure as part of training, and show that it consistently improves the quality of ML models compared to traditional choices of energy reference, for different classes of materials and different model architectures. On a practical level, we trace the improved performance to the ability of this self-aligning scheme to match the most prominent features in the DOS. More broadly, we believe that this work highlights the importance of incorporating insights into the nature of the physical target into the definition of the architecture and of the appropriate figures of merit for machine-learning models, that translate in better transferability and overall performance. This record contains all the necessary data files and scripts to support the results presented in the paper with the same title.

Materials Cloud sections using this data

No Explore or Discover sections associated with this archive record.

Files

File name Size Description
dataset.tar.gz
MD5md5:35bc9589b6102e93910f16e7db607054
5.0 GiB Dataset and scripts necessary to reproduce the results of the paper

License

Files and data are licensed under the terms of the following license: Creative Commons Attribution 4.0 International.
Metadata, except for email addresses, are licensed under the Creative Commons Attribution Share-Alike 4.0 International license.

Keywords

ERC MARVEL SNSF machine learning electronic structure density of states EPFL