Hide metadata

dc.date.accessioned2019-09-11T05:24:38Z
dc.date.available2019-09-11T05:24:38Z
dc.date.created2019-01-17T22:45:08Z
dc.date.issued2018
dc.identifier.citationAgrell, Christian Eldevik, Simen Hafver, Andreas Pedersen, Frank Børre Stensrud, Erik Huseby, Arne . Pitfalls of machine learning for tail events in high risk environments. Safety and Reliability – Safe Societies in a Changing World. Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway. 2018, 3043-3051 Taylor & Francis
dc.identifier.urihttp://hdl.handle.net/10852/70337
dc.description.abstractMost of today’s Machine Learning (ML) methods and implementations are based on correlations, in the sense of a statistical relationship between a set of inputs and the output(s) under inves- tigation. The relationship might be obscure to the human mind, but through the use of ML, mathematics and statistics makes it seemingly apparent. However, to base safety critical decisions on such methods suffer from the same pitfalls as decisions based on any other correlation metric that disregards causality. Causality is key to ensure that applied mitigation tactics will actually affect the outcome in the desired way. This paper reviews the current situation and challenges of applying ML in high risk environments. It further outlines how phenomenological knowledge, together with an uncertainty-based risk perspective can be incorporated to alleviate the missing causality considerations in current practice.
dc.description.abstractPitfalls of machine learning for tail events in high risk environments
dc.languageEN
dc.publisherTaylor & Francis
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.titlePitfalls of machine learning for tail events in high risk environments
dc.typeChapter
dc.creator.authorAgrell, Christian
dc.creator.authorEldevik, Simen
dc.creator.authorHafver, Andreas
dc.creator.authorPedersen, Frank Børre
dc.creator.authorStensrud, Erik
dc.creator.authorHuseby, Arne
cristin.unitcode185,15,13,35
cristin.unitnameStokastisk analyse, finans, forsikring og risiko
cristin.ispublishedtrue
cristin.fulltextoriginal
dc.identifier.cristin1659819
dc.identifier.bibliographiccitationinfo:ofi/fmt:kev:mtx:ctx&ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.btitle=Safety and Reliability – Safe Societies in a Changing World. Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway&rft.spage=3043&rft.date=2018
dc.identifier.startpage3043
dc.identifier.endpage3051
dc.identifier.pagecount3234
dc.identifier.doihttps://doi.org/10.1201/9781351174664
dc.identifier.urnURN:NBN:no-73474
dc.type.documentBokkapittel
dc.type.peerreviewedPeer reviewed
dc.source.isbn9781351174657
dc.identifier.fulltextFulltext https://www.duo.uio.no/bitstream/handle/10852/70337/1/ESREL%2B2018%2B-%2BPitfalls%2Bof%2Bmachine%2Blearning%2Bfor%2Btail%2Bevents%2Bin%2Bhigh%2Brisk%2Benvironments.pdf
dc.type.versionPublishedVersion
cristin.btitleSafety and Reliability – Safe Societies in a Changing World. Proceedings of ESREL 2018, June 17-21, 2018, Trondheim, Norway
dc.relation.projectNFR/276282


Files in this item

Appears in the following Collection

Hide metadata

Attribution-NonCommercial-NoDerivatives 4.0 International
This item's license is: Attribution-NonCommercial-NoDerivatives 4.0 International