Identifying domains of applicability of machine learning models for materials science

Christopher Sutton (), Mario Boley (), Luca M. Ghiringhelli (), Matthias Rupp, Jilles Vreeken and Matthias Scheffler
Additional contact information
Christopher Sutton: Fritz Haber Institute of the Max Planck Society
Mario Boley: Monash University
Luca M. Ghiringhelli: Fritz Haber Institute of the Max Planck Society
Matthias Rupp: Fritz Haber Institute of the Max Planck Society
Jilles Vreeken: CISPA Helmholtz Center for Information Security
Matthias Scheffler: Fritz Haber Institute of the Max Planck Society

Nature Communications, 2020, vol. 11, issue 1, 1-9

Abstract: Abstract 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.

Date: 2020
References: Add references at CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.nature.com/articles/s41467-020-17112-9 Abstract (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nat:natcom:v:11:y:2020:i:1:d:10.1038_s41467-020-17112-9

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-020-17112-9

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().