CISPA
Browse
cispa_all_3252.pdf (5.34 MB)

Identifying Domains of Applicability of Machine Learning Models for Materials Science

Download (5.34 MB)
journal contribution
posted on 2023-11-29, 18:07 authored by Christopher Sutton, Mario Boley, Luca M. Ghiringhelli, Matthias Rupp, Jilles VreekenJilles Vreeken, Matthias Scheffler
Although machine learning (ML) models promise to substantially accelerate the discovery of novel materials, their performance is often still insufficient to draw reliable conclusions. Improved ML models are therefore actively researched, but their design is currently guided mainly by monitoring the average model test error. This can render different models indistinguishable although their performance differs substantially across materials, or it can make a model appear generally insufficient while it actually works well in specific sub-domains. Here, we present a method, based on subgroup discovery, for detecting domains of applicability (DA) of models within a materials class. The utility of this approach is demonstrated by analyzing three state-of-the-art ML models for predicting the formation energy of transparent conducting oxides. We find that, despite having a mutually indistinguishable and unsatisfactory average error, the models have DAs with distinctive features and notably improved performance.

History

Preferred Citation

Christopher Sutton, Mario Boley, Luca Ghiringhelli, Matthias Rupp, Jilles Vreeken and Matthias Scheffler. Identifying Domains of Applicability of Machine Learning Models for Materials Science. In: Nature Communications. 2020.

Primary Research Area

  • Trustworthy Information Processing

Legacy Posted Date

2020-10-15

Journal

Nature Communications

Open Access Type

  • Gold

Sub Type

  • Article

BibTeX

@article{cispa_all_3252, title = "Identifying Domains of Applicability of Machine Learning Models for Materials Science", author = "Sutton, Christopher and Boley, Mario and Ghiringhelli, Luca M. and Rupp, Matthias and Vreeken, Jilles and Scheffler, Matthias", journal="{Nature Communications}", year="2020", }

Usage metrics

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC