Unmasking Clever Hans predictors and assessing what machines really learn

Nonlinear machine learning methods have good predictive ability but the lack of transparency of the algorithms can limit their use. Here the authors investigate how these methods approach learning in order to assess the dependability of their decision making.

Guardado en:
Detalles Bibliográficos
Autores principales: Sebastian Lapuschkin, Stephan Wäldchen, Alexander Binder, Grégoire Montavon, Wojciech Samek, Klaus-Robert Müller
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2019
Materias:
Q
Acceso en línea:https://doaj.org/article/ae85cf785d3645e68f9b5bfe27287566
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:ae85cf785d3645e68f9b5bfe27287566
record_format dspace
spelling oai:doaj.org-article:ae85cf785d3645e68f9b5bfe272875662021-12-02T16:57:35ZUnmasking Clever Hans predictors and assessing what machines really learn10.1038/s41467-019-08987-42041-1723https://doaj.org/article/ae85cf785d3645e68f9b5bfe272875662019-03-01T00:00:00Zhttps://doi.org/10.1038/s41467-019-08987-4https://doaj.org/toc/2041-1723Nonlinear machine learning methods have good predictive ability but the lack of transparency of the algorithms can limit their use. Here the authors investigate how these methods approach learning in order to assess the dependability of their decision making.Sebastian LapuschkinStephan WäldchenAlexander BinderGrégoire MontavonWojciech SamekKlaus-Robert MüllerNature PortfolioarticleScienceQENNature Communications, Vol 10, Iss 1, Pp 1-8 (2019)
institution DOAJ
collection DOAJ
language EN
topic Science
Q
spellingShingle Science
Q
Sebastian Lapuschkin
Stephan Wäldchen
Alexander Binder
Grégoire Montavon
Wojciech Samek
Klaus-Robert Müller
Unmasking Clever Hans predictors and assessing what machines really learn
description Nonlinear machine learning methods have good predictive ability but the lack of transparency of the algorithms can limit their use. Here the authors investigate how these methods approach learning in order to assess the dependability of their decision making.
format article
author Sebastian Lapuschkin
Stephan Wäldchen
Alexander Binder
Grégoire Montavon
Wojciech Samek
Klaus-Robert Müller
author_facet Sebastian Lapuschkin
Stephan Wäldchen
Alexander Binder
Grégoire Montavon
Wojciech Samek
Klaus-Robert Müller
author_sort Sebastian Lapuschkin
title Unmasking Clever Hans predictors and assessing what machines really learn
title_short Unmasking Clever Hans predictors and assessing what machines really learn
title_full Unmasking Clever Hans predictors and assessing what machines really learn
title_fullStr Unmasking Clever Hans predictors and assessing what machines really learn
title_full_unstemmed Unmasking Clever Hans predictors and assessing what machines really learn
title_sort unmasking clever hans predictors and assessing what machines really learn
publisher Nature Portfolio
publishDate 2019
url https://doaj.org/article/ae85cf785d3645e68f9b5bfe27287566
work_keys_str_mv AT sebastianlapuschkin unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
AT stephanwaldchen unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
AT alexanderbinder unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
AT gregoiremontavon unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
AT wojciechsamek unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
AT klausrobertmuller unmaskingcleverhanspredictorsandassessingwhatmachinesreallylearn
_version_ 1718382522294861824