Simple view
Full metadata view
Authors
Statistics
Processing of missing data by neural networks
2018
book section
conference proceedings
Author
Śmieja Marek
Struski Łukasz
Tabor Jacek
Zieliński Bartosz
Spurek Przemysław
Book title / Journal title
Advances in neural information processing systems 31 : Annual Conference on Neural Information Processing Systems 2018
publisher
Saarbrücken : Schloss Dagstuhl Leibniz Center for Informatics
Pages
2724-2734
Date accessed
2019-09-25
Language
English
Book language / Journal language
English
Abstract in English
We propose a general, theoretically justified mechanism for processing missing data by neural networks. Our idea is to replace typical neuron's response in the first hidden layer by its expected value. This approach can be applied for various types of networks at minimal cost in their modification. Moreover, in contrast to recent approaches, it does not require complete data for training. Experimental results performed on different types of architectures show that our method gives better results than typical imputation strategies and other methods dedicated for incomplete data.
Conference
Advances in Neural Information Processing Systems 31 (NeurIPS 2018)
Conference short name
NIPS
Conference start date
2018-12-03
End date conference
2018-12-08
Conference city
Montreal
Conference country
Kanada
Conference type
international
Affiliation
Wydział Matematyki i Informatyki : Instytut Informatyki i Matematyki KomputerowejWydział Matematyki i Informatyki : Instytut Matematyki
dc.abstract.en | We propose a general, theoretically justified mechanism for processing missing data by neural networks. Our idea is to replace typical neuron's response in the first hidden layer by its expected value. This approach can be applied for various types of networks at minimal cost in their modification. Moreover, in contrast to recent approaches, it does not require complete data for training. Experimental results performed on different types of architectures show that our method gives better results than typical imputation strategies and other methods dedicated for incomplete data. | pl |
dc.affiliation | Wydział Matematyki i Informatyki : Instytut Informatyki i Matematyki Komputerowej | pl |
dc.affiliation | Wydział Matematyki i Informatyki : Instytut Matematyki | pl |
dc.conference | Advances in Neural Information Processing Systems 31 (NeurIPS 2018) | pl |
dc.conference.city | Montreal | |
dc.conference.country | Kanada | |
dc.conference.datefinish | 2018-12-08 | |
dc.conference.datestart | 2018-12-03 | |
dc.conference.shortcut | NIPS | |
dc.contributor.author | Śmieja, Marek - 135996 | pl |
dc.contributor.author | Struski, Łukasz - 135994 | pl |
dc.contributor.author | Tabor, Jacek - 132362 | pl |
dc.contributor.author | Zieliński, Bartosz - 106948 | pl |
dc.contributor.author | Spurek, Przemysław - 135993 | pl |
dc.date.accession | 2019-09-25 | pl |
dc.date.accessioned | 2019-09-25T07:43:43Z | |
dc.date.available | 2019-09-25T07:43:43Z | |
dc.date.issued | 2018 | pl |
dc.date.openaccess | 0 | |
dc.description.accesstime | w momencie opublikowania | |
dc.description.conftype | international | pl |
dc.description.physical | 2724-2734 | pl |
dc.description.publication | 1,12 | pl |
dc.description.version | ostateczna wersja wydawcy | |
dc.identifier.project | ROD UJ / O | pl |
dc.identifier.uri | https://ruj.uj.edu.pl/xmlui/handle/item/83349 | |
dc.identifier.weblink | https://papers.nips.cc/paper/7537-processing-of-missing-data-by-neural-networks.pdf | pl |
dc.language | eng | pl |
dc.language.container | eng | pl |
dc.pubinfo | Saarbrücken : Schloss Dagstuhl Leibniz Center for Informatics | pl |
dc.rights | Dodaję tylko opis bibliograficzny | * |
dc.rights.licence | OTHER | |
dc.rights.uri | * | |
dc.share.type | inne | |
dc.subtype | ConferenceProceedings | pl |
dc.title | Processing of missing data by neural networks | pl |
dc.title.container | Advances in neural information processing systems 31 : Annual Conference on Neural Information Processing Systems 2018 | pl |
dc.type | BookSection | pl |
dspace.entity.type | Publication |
dc.abstract.enpl
We propose a general, theoretically justified mechanism for processing missing data by neural networks. Our idea is to replace typical neuron's response in the first hidden layer by its expected value. This approach can be applied for various types of networks at minimal cost in their modification. Moreover, in contrast to recent approaches, it does not require complete data for training. Experimental results performed on different types of architectures show that our method gives better results than typical imputation strategies and other methods dedicated for incomplete data. dc.affiliationpl
Wydział Matematyki i Informatyki : Instytut Informatyki i Matematyki Komputerowej dc.affiliationpl
Wydział Matematyki i Informatyki : Instytut Matematyki dc.conferencepl
Advances in Neural Information Processing Systems 31 (NeurIPS 2018) dc.conference.city
Montreal dc.conference.country
Kanada dc.conference.datefinish
2018-12-08 dc.conference.datestart
2018-12-03 dc.conference.shortcut
NIPS dc.contributor.authorpl
Śmieja, Marek - 135996 dc.contributor.authorpl
Struski, Łukasz - 135994 dc.contributor.authorpl
Tabor, Jacek - 132362 dc.contributor.authorpl
Zieliński, Bartosz - 106948 dc.contributor.authorpl
Spurek, Przemysław - 135993 dc.date.accessionpl
2019-09-25 dc.date.accessioned
2019-09-25T07:43:43Z dc.date.available
2019-09-25T07:43:43Z dc.date.issuedpl
2018 dc.date.openaccess
0 dc.description.accesstime
w momencie opublikowania dc.description.conftypepl
international dc.description.physicalpl
2724-2734 dc.description.publicationpl
1,12 dc.description.version
ostateczna wersja wydawcy dc.identifier.projectpl
ROD UJ / O dc.identifier.uri
https://ruj.uj.edu.pl/xmlui/handle/item/83349 dc.identifier.weblinkpl
https://papers.nips.cc/paper/7537-processing-of-missing-data-by-neural-networks.pdf dc.languagepl
eng dc.language.containerpl
eng dc.pubinfopl
Saarbrücken : Schloss Dagstuhl Leibniz Center for Informatics dc.rights*
Dodaję tylko opis bibliograficzny dc.rights.licence
OTHER dc.rights.uri*
dc.share.type
inne dc.subtypepl
ConferenceProceedings dc.titlepl
Processing of missing data by neural networks dc.title.containerpl
Advances in neural information processing systems 31 : Annual Conference on Neural Information Processing Systems 2018 dc.typepl
BookSection dspace.entity.type
Publication Affiliations
Wydział Matematyki i Informatyki
Śmieja, Marek
Struski, Łukasz
Tabor, Jacek
Zieliński, Bartosz
Spurek, Przemysław