Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic...
Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials
About this item
Full title
Author / Creator
Publisher
Bristol: IOP Publishing
Journal title
Language
English
Formats
Publication information
Publisher
Bristol: IOP Publishing
Subjects
More information
Scope and Contents
Contents
Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established
a-posteriori
UQ methods, including ensemble methods, the dropout method, the delta method,...
Alternative Titles
Full title
Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2756718627
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2756718627
Other Identifiers
ISSN
2632-2153
E-ISSN
2632-2153
DOI
10.1088/2632-2153/aca7b1
