On the pitfalls of Gaussian likelihood scoring for causal discovery
dc.contributor.author
Schultheiss, Christoph
dc.contributor.author
Bühlmann, Peter
dc.date.accessioned
2023-05-26T05:15:55Z
dc.date.available
2023-05-26T03:22:53Z
dc.date.available
2023-05-26T05:15:55Z
dc.date.issued
2023-01
dc.identifier.issn
2193-3685
dc.identifier.issn
2193-3677
dc.identifier.other
10.1515/jci-2022-0068
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/613690
dc.identifier.doi
10.3929/ethz-b-000613690
dc.description.abstract
We consider likelihood score-based methods for causal discovery in structural causal models. In particular, we focus on Gaussian scoring and analyze the effect of model misspecification in terms of non-Gaussian error distribution. We present a surprising negative result for Gaussian likelihood scoring in combination with nonparametric regression methods.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
De Gruyter
en_US
dc.rights.uri
http://creativecommons.org/licenses/by/4.0/
dc.subject
graphical models
en_US
dc.subject
model misspecification
en_US
dc.subject
non-parametric regression
en_US
dc.subject
structural causal models
en_US
dc.title
On the pitfalls of Gaussian likelihood scoring for causal discovery
en_US
dc.type
Journal Article
dc.rights.license
Creative Commons Attribution 4.0 International
dc.date.published
2023-05-11
ethz.journal.title
Journal of Causal Inference
ethz.journal.volume
11
en_US
ethz.journal.issue
1
en_US
ethz.pages.start
20220068
en_US
ethz.size
11 p.
en_US
ethz.version.deposit
publishedVersion
en_US
ethz.grant
Statistics, Prediction and Causality for Large-Scale Data
en_US
ethz.identifier.wos
ethz.identifier.scopus
ethz.identifier.nebis
ethz.publication.place
Berlin
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02000 - Dep. Mathematik / Dep. of Mathematics::02537 - Seminar für Statistik (SfS) / Seminar for Statistics (SfS)::03502 - Bühlmann, Peter L. / Bühlmann, Peter L.
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02000 - Dep. Mathematik / Dep. of Mathematics::02537 - Seminar für Statistik (SfS) / Seminar for Statistics (SfS)::03502 - Bühlmann, Peter L. / Bühlmann, Peter L.
ethz.grant.agreementno
786461
ethz.grant.fundername
EC
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.program
H2020
ethz.date.deposited
2023-05-26T03:22:54Z
ethz.source
SCOPUS
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2024-02-02T23:45:15Z
ethz.rosetta.lastUpdated
2024-02-02T23:45:15Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=On%20the%20pitfalls%20of%20Gaussian%20likelihood%20scoring%20for%20causal%20discovery&rft.jtitle=Journal%20of%20Causal%20Inference&rft.date=2023-01&rft.volume=11&rft.issue=1&rft.spage=20220068&rft.issn=2193-3685&2193-3677&rft.au=Schultheiss,%20Christoph&B%C3%BChlmann,%20Peter&rft.genre=article&rft_id=info:doi/10.1515/jci-2022-0068&
Files in this item
Publication type
-
Journal Article [132162]