Stream-based active learning for sliding windows under the influence of verification latency
dc.date.accessioned | 2022-08-16T09:17:44Z | |
dc.date.available | 2022-08-16T09:17:44Z | |
dc.date.issued | 2021-11-18 | |
dc.identifier | doi:10.17170/kobra-202206016277 | |
dc.identifier.uri | http://hdl.handle.net/123456789/14057 | |
dc.description.sponsorship | Gefördert im Rahmen des Projekts DEAL | ger |
dc.language.iso | eng | |
dc.rights | Namensnennung 4.0 International | * |
dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | * |
dc.subject | classification | eng |
dc.subject | active learning | eng |
dc.subject | evolving data streams | eng |
dc.subject | concept drift | eng |
dc.subject | verification latency | eng |
dc.subject | label delay | eng |
dc.subject.ddc | 004 | |
dc.title | Stream-based active learning for sliding windows under the influence of verification latency | eng |
dc.type | Aufsatz | |
dcterms.abstract | Stream-based active learning (AL) strategies minimize the labeling effort by querying labels that improve the classifier’s performance the most. So far, these strategies neglect the fact that an oracle or expert requires time to provide a queried label. We show that existing AL methods deteriorate or even fail under the influence of such verification latency. The problem with these methods is that they estimate a label’s utility on the currently available labeled data. However, when this label would arrive, some of the current data may have gotten outdated and new labels have arrived. In this article, we propose to simulate the available data at the time when the label would arrive. Therefore, our method Forgetting and Simulating (FS) forgets outdated information and simulates the delayed labels to get more realistic utility estimates. We assume to know the label’s arrival date a priori and the classifier’s training data to be bounded by a sliding window. Our extensive experiments show that FS improves stream-based AL strategies in settings with both, constant and variable verification latency. | eng |
dcterms.accessRights | open access | |
dcterms.creator | Pham, Tuan | |
dcterms.creator | Kottke, Daniel | |
dcterms.creator | Krempl, Georg | |
dcterms.creator | Sick, Bernhard | |
dc.relation.doi | doi:10.1007/s10994-021-06099-z | |
dc.subject.swd | Aktives Maschinelles Lernen | ger |
dc.subject.swd | Datenstrom | ger |
dc.subject.swd | Klassifikation | ger |
dc.subject.swd | Simulation | ger |
dc.subject.swd | Etikettieren | ger |
dc.subject.swd | Algorithmus | ger |
dc.type.version | publishedVersion | |
dcterms.source.identifier | eissn:1573-0565 | |
dcterms.source.issue | Issue 6 | |
dcterms.source.journal | Machine Learning | eng |
dcterms.source.pageinfo | 2011-2036 | |
dcterms.source.volume | Volume 111 | |
kup.iskup | false |
Dateien zu dieser Ressource
Das Dokument erscheint in:
-
Artikel [1223]