Why big data and compute are not necessarily the path to big materials science
Machine learning is an increasingly important tool for materials science. Here, the authors suggest that its contextual use, including careful assessment of resources and bias, judicious model selection, and an understanding of its limitations, will help researchers to expedite scientific discovery.
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-08-01
|
| Series: | Communications Materials |
| Online Access: | https://doi.org/10.1038/s43246-022-00283-x |
| _version_ | 1798032436669448192 |
|---|---|
| author | Naohiro Fujinuma Brian DeCost Jason Hattrick-Simpers Samuel E. Lofland |
| author_facet | Naohiro Fujinuma Brian DeCost Jason Hattrick-Simpers Samuel E. Lofland |
| author_sort | Naohiro Fujinuma |
| collection | DOAJ |
| description | Machine learning is an increasingly important tool for materials science. Here, the authors suggest that its contextual use, including careful assessment of resources and bias, judicious model selection, and an understanding of its limitations, will help researchers to expedite scientific discovery. |
| first_indexed | 2024-04-11T20:12:44Z |
| format | Article |
| id | doaj.art-92ea40334d394a9784760a26b7c35f56 |
| institution | Directory Open Access Journal |
| issn | 2662-4443 |
| language | English |
| last_indexed | 2024-04-11T20:12:44Z |
| publishDate | 2022-08-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Materials |
| spelling | doaj.art-92ea40334d394a9784760a26b7c35f562022-12-22T04:05:03ZengNature PortfolioCommunications Materials2662-44432022-08-01311910.1038/s43246-022-00283-xWhy big data and compute are not necessarily the path to big materials scienceNaohiro Fujinuma0Brian DeCost1Jason Hattrick-Simpers2Samuel E. Lofland3Department of Chemical Engineering, Rowan UniversityMaterial Measurement Laboratory, National Institute of Standards and TechnologyDepartment of Materials Science and Engineering, University of TorontoDepartment of Physics and Astronomy, Rowan UniversityMachine learning is an increasingly important tool for materials science. Here, the authors suggest that its contextual use, including careful assessment of resources and bias, judicious model selection, and an understanding of its limitations, will help researchers to expedite scientific discovery.https://doi.org/10.1038/s43246-022-00283-x |
| spellingShingle | Naohiro Fujinuma Brian DeCost Jason Hattrick-Simpers Samuel E. Lofland Why big data and compute are not necessarily the path to big materials science Communications Materials |
| title | Why big data and compute are not necessarily the path to big materials science |
| title_full | Why big data and compute are not necessarily the path to big materials science |
| title_fullStr | Why big data and compute are not necessarily the path to big materials science |
| title_full_unstemmed | Why big data and compute are not necessarily the path to big materials science |
| title_short | Why big data and compute are not necessarily the path to big materials science |
| title_sort | why big data and compute are not necessarily the path to big materials science |
| url | https://doi.org/10.1038/s43246-022-00283-x |
| work_keys_str_mv | AT naohirofujinuma whybigdataandcomputearenotnecessarilythepathtobigmaterialsscience AT briandecost whybigdataandcomputearenotnecessarilythepathtobigmaterialsscience AT jasonhattricksimpers whybigdataandcomputearenotnecessarilythepathtobigmaterialsscience AT samuelelofland whybigdataandcomputearenotnecessarilythepathtobigmaterialsscience |