DScribe: Library of Descriptors for Machine Learning in Materials Science

By Lauri Himanen, M. Jäger, Eiaki V. Morooka, Filippo Federici Canova, Y. S. Ranawat, D. Gao, Patrick Rinke, Adam S. Foster. Computer Physics Communications, 2019

748 citations recorded by the source corpus.

DScribe is a software package for machine learning that provides popular feature transformations ("descriptors") for atomistic materials simulations. DScribe accelerates the application of machine learning for atomistic property prediction by providing user-friendly, off-the-shelf descriptor implementations. The package currently contains implementations for Coulomb matrix, Ewald sum matrix, sine matrix, Many-body Tensor Representation (MBTR), Atom-centered Symmetry Function (ACSF) and Smooth Overlap of Atomic Positions (SOAP). Usage of the package is illustrated for two different applications: formation energy prediction for solids and ionic charge prediction for atoms in organic molecules. The package is freely available under the open-source Apache License 2.0.

Subfields: Computational Theory and Mathematics, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/93e4cb65cb5bfa568e57495c804c4c5e2324ccaf

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

1. Shyue Ping Ong
2. Chi Chen
3. Ryan Jacobs
4. Dane Morgan
5. Alok N. Choudhary
6. Chris Wolverton

In Materials Chemistry

1. Shyue Ping Ong
2. Ryan Jacobs
3. Dane Morgan
4. Chi Chen
5. Anubhav Jain
6. Alok N. Choudhary
7. Chris Wolverton
8. Simon J. L. Billinge

Citations in the corpus

• Modeling the formation and growth of atmospheric molecular clusters: A review (2020, methodology): “Some of them have been implemented as libraries in Python such as Quantum Machine Learning (QML) (Christensen et al., 2017) and DScribe (Himanen et al., 2020).”
• Recent advances and applications of deep learning methods in materials science (2021, methodology): “There have been libraries of such descriptors developed such as MatMiner [161] and DScribe [168].”
• Opportunities and Challenges for Machine Learning in Materials Science (2020, background): “…et al. 2019 207 Materials informatics Agrawal and Choudhary 2019 208 Deep materials informatics: Applications of deep learning in materials science Himanen et al. 2019 209 Data-driven materials science: status, challenges and perspectives Reyes and Maruyama 2019 210 The machine learning…”