Putting Chemical Knowledge to Work in Machine Learning for Reactivity
dc.contributor.author
Jorner, Kjell
dc.date.accessioned
2023-03-13T07:41:01Z
dc.date.available
2023-03-10T16:20:26Z
dc.date.available
2023-03-13T07:41:01Z
dc.date.issued
2023-02-22
dc.identifier.issn
0009-4293
dc.identifier.other
10.2533/chimia.2023.22
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/602723
dc.identifier.doi
10.3929/ethz-b-000602723
dc.description.abstract
Machine learning has been used to study chemical reactivity for a long time in fields such as physical organic chemistry, chemometrics and cheminformatics. Recent advances in computer science have resulted in deep neural networks that can learn directly from the molecular structure. Neural networks are a good choice when large amounts of data are available. However, many datasets in chemistry are small, and models utilizing chemical knowledge are required for good performance. Adding chemical knowledge can be achieved either by adding more information about the molecules or by adjusting the model architecture itself. The current method of choice for adding more information is descriptors based on computed quantum-chemical properties. Exciting new research directions show that it is possible to augment deep learning with such descriptors for better per formance in the low-data regime. To modify the models, differentiable programming enables seamless merging of neural networks with mathematical models from chemistry and physics. The resulting methods are also more data-efficient and make better predictions for molecules that are different from the initial dataset on which they were trained. Application of these chemistry-informed machine learning methods promise to accelerate research in fields such as drug design, materials design, catalysis and reactivity.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
Swiss Chemical Society
dc.rights.uri
http://creativecommons.org/licenses/by/4.0/
dc.subject
Digital chemistry
en_US
dc.subject
Machine learning
en_US
dc.subject
Reactivity
en_US
dc.title
Putting Chemical Knowledge to Work in Machine Learning for Reactivity
en_US
dc.type
Journal Article
dc.rights.license
Creative Commons Attribution 4.0 International
ethz.journal.title
Chimia
ethz.journal.volume
77
en_US
ethz.journal.issue
1/2
en_US
ethz.journal.abbreviated
Chimia
ethz.pages.start
22
en_US
ethz.pages.end
30
en_US
ethz.version.deposit
publishedVersion
en_US
ethz.publication.place
Bern
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02020 - Dep. Chemie und Angewandte Biowiss. / Dep. of Chemistry and Applied Biosc.::02516 - Inst. f. Chemie- und Bioingenieurwiss. / Inst. Chemical and Bioengineering::09781 - Jorner, Kjell / Jorner, Kjell
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02020 - Dep. Chemie und Angewandte Biowiss. / Dep. of Chemistry and Applied Biosc.::02516 - Inst. f. Chemie- und Bioingenieurwiss. / Inst. Chemical and Bioengineering::09781 - Jorner, Kjell / Jorner, Kjell
en_US
ethz.date.deposited
2023-03-10T16:20:26Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2023-03-13T07:41:02Z
ethz.rosetta.lastUpdated
2024-02-02T20:57:09Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Putting%20Chemical%20Knowledge%20to%20Work%20in%20Machine%20Learning%20for%20Reactivity&rft.jtitle=Chimia&rft.date=2023-02-22&rft.volume=77&rft.issue=1/2&rft.spage=22&rft.epage=30&rft.issn=0009-4293&rft.au=Jorner,%20Kjell&rft.genre=article&rft_id=info:doi/10.2533/chimia.2023.22&
Files in this item
Publication type
-
Journal Article [130260]