Assessing the frontier : Active learning, model accuracy, and multi-objective candidate discovery and optimization

Del Rosario, Zachary; Rupp, Matthias; Kim, Yoolhee; Antono, Erin; Ling, Julia

Assessing the frontier : Active learning, model accuracy, and multi-objective candidate discovery and optimization

Dateien

Rosario_2-1rkaqygcxvcal5.pdfGröße: 3.05 MBDownloads: 431

Datum

2020

Autor:innen

Del Rosario, Zachary

Rupp, Matthias

Kim, Yoolhee

Antono, Erin

Ling, Julia

Open Access-Veröffentlichung

Open Access Hybrid

Sammlungen

Informatik und Informationswissenschaft

Publikationstyp

Zeitschriftenartikel

Publikationsstatus

Published

Erschienen in

The Journal of Chemical Physics. American Institute of Physics (AIP). 2020, 153(2), 024112. ISSN 0021-9606. eISSN 1089-7690. Available under: doi: 10.1063/5.0006124

Zusammenfassung

Discovering novel chemicals and materials can be greatly accelerated by iterative machine learning-informed proposal of candidates-active learning. However, standard global error metrics for model quality are not predictive of discovery performance and can be misleading. We introduce the notion of Pareto shell error to help judge the suitability of a model for proposing candidates. Furthermore, through synthetic cases, an experimental thermoelectric dataset and a computational organic molecule dataset, we probe the relation between acquisition function fidelity and active learning performance. Results suggest novel diagnostic tools, as well as new insights for the acquisition function design.

Fachgebiet (DDC)

004 Informatik

Zitieren

ISO 690

DEL ROSARIO, Zachary, Matthias RUPP, Yoolhee KIM, Erin ANTONO, Julia LING, 2020. Assessing the frontier : Active learning, model accuracy, and multi-objective candidate discovery and optimization. In: The Journal of Chemical Physics. American Institute of Physics (AIP). 2020, 153(2), 024112. ISSN 0021-9606. eISSN 1089-7690. Available under: doi: 10.1063/5.0006124

BibTex

@article{DelRosario2020-07-14Asses-52563,
  year={2020},
  doi={10.1063/5.0006124},
  title={Assessing the frontier : Active learning, model accuracy, and multi-objective candidate discovery and optimization},
  number={2},
  volume={153},
  issn={0021-9606},
  journal={The Journal of Chemical Physics},
  author={Del Rosario, Zachary and Rupp, Matthias and Kim, Yoolhee and Antono, Erin and Ling, Julia},
  note={Article Number: 024112}
}

RDF

<rdf:RDF
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:bibo="http://purl.org/ontology/bibo/"
    xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#"
    xmlns:foaf="http://xmlns.com/foaf/0.1/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > 
  <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/52563">
    <dcterms:hasPart rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/52563/1/Rosario_2-1rkaqygcxvcal5.pdf"/>
    <dc:contributor>Ling, Julia</dc:contributor>
    <dcterms:abstract xml:lang="eng">Discovering novel chemicals and materials can be greatly accelerated by iterative machine learning-informed proposal of candidates-active learning. However, standard global error metrics for model quality are not predictive of discovery performance and can be misleading. We introduce the notion of Pareto shell error to help judge the suitability of a model for proposing candidates. Furthermore, through synthetic cases, an experimental thermoelectric dataset and a computational organic molecule dataset, we probe the relation between acquisition function fidelity and active learning performance. Results suggest novel diagnostic tools, as well as new insights for the acquisition function design.</dcterms:abstract>
    <dcterms:rights rdf:resource="http://creativecommons.org/licenses/by/4.0/"/>
    <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/>
    <dc:contributor>Rupp, Matthias</dc:contributor>
    <dspace:hasBitstream rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/52563/1/Rosario_2-1rkaqygcxvcal5.pdf"/>
    <foaf:homepage rdf:resource="http://localhost:8080/"/>
    <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2021-01-26T10:11:09Z</dcterms:available>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dc:language>eng</dc:language>
    <dc:rights>Attribution 4.0 International</dc:rights>
    <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/52563"/>
    <dc:creator>Antono, Erin</dc:creator>
    <dc:creator>Rupp, Matthias</dc:creator>
    <dcterms:issued>2020-07-14</dcterms:issued>
    <dcterms:title>Assessing the frontier : Active learning, model accuracy, and multi-objective candidate discovery and optimization</dcterms:title>
    <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2021-01-26T10:11:09Z</dc:date>
    <dc:contributor>Antono, Erin</dc:contributor>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dc:contributor>Del Rosario, Zachary</dc:contributor>
    <dc:creator>Del Rosario, Zachary</dc:creator>
    <dc:creator>Ling, Julia</dc:creator>
    <dc:creator>Kim, Yoolhee</dc:creator>
    <dc:contributor>Kim, Yoolhee</dc:contributor>
  </rdf:Description>
</rdf:RDF>

Universitätsbibliographie

Ja

Begutachtet

Ja