Stacked-MLkNN: A stacking based improvement to Multi-Label k-Nearest Neighbours

DC FieldValueLanguage
dc.contributor.authorPakrashi, Arjun-
dc.contributor.authorMacNamee, Brian-
dc.contributor.editorTorgo, Luís-
dc.contributor.editorKrawczyk, Bartosz-
dc.contributor.editorBranco, Paula-
dc.contributor.editorMoniz, Nuno-
dc.date.accessioned2019-04-18T10:10:01Z-
dc.date.available2019-04-18T10:10:01Z-
dc.date.copyright2017 the Authorsen_US
dc.date.issued2017-09-22-
dc.identifier.urihttp://hdl.handle.net/10197/10041-
dc.descriptionThe 1st International Workshop on Learning with Imbalanced Domains: Theory and Applications (LIDTA 2017), Skopje, Macedonia, 18-22 Septemberen_US
dc.description.abstractMulti-label classification deals with problems where each datapoint can be assigned to more than one class, or label, at the same time. The simplest approach for such problems is to train independent binary classification models for each label and use these models to independently predict a set of relevant labels for a datapoint. MLkNN is an instance-based lazy learning algorithm for multi-label classification that takes this approach. MLkNN, and similar algorithms, however, do not exploit associations which may exist between the set of potential labels. These methods also suffer from imbalance in the frequency of labels in a training dataset. This work attempts to improve the predictions of MLkNN by implementing a two-layer stack-like method, Stacked-MLkNN which exploits the label associations. Experiments show that Stacked-MLkNN produces better predictions than MLkNN and several other state-of-the-art instance-based learning algorithms.en_US
dc.description.sponsorshipScience Foundation Irelanden_US
dc.language.isoenen_US
dc.publisherJMLRen_US
dc.relation.ispartofProceedings of Machine Learning Research. Volume 74: First International Workshop on Learning with Imbalanced Domains: Theory and Applications, 22 September 2017, ECML-PKDD, Skopje, Macedoniaen_US
dc.subjectMulti-labelen_US
dc.subjectStackingen_US
dc.subjectInstance-based learningen_US
dc.titleStacked-MLkNN: A stacking based improvement to Multi-Label k-Nearest Neighboursen_US
dc.typeConference Publicationen_US
dc.internal.authorcontactotherbrian.macnamee@ucd.ieen_US
dc.internal.webversionshttp://lidta.dcc.fc.up.pt/2017/index.html-
dc.statusNot peer revieweden_US
dc.neeo.contributorPakrashi|Arjun|aut|-
dc.neeo.contributorMacNamee|Brian|aut|-
dc.neeo.contributorTorgo|Luís|edt|-
dc.neeo.contributorKrawczyk|Bartosz|edt|-
dc.neeo.contributorBranco|Paula|edt|-
dc.neeo.contributorMoniz|Nuno|edt|-
dc.date.updated2018-07-06T15:10:01Z-
dc.relation.datasethttp://proceedings.mlr.press/v74/en_US
dc.identifier.grantidSFI/12/RC/2289-
item.fulltextWith Fulltext-
item.grantfulltextopen-
Appears in Collections:Computer Science Research Collection
Files in This Item:
File Description SizeFormat 
pakrashi17a.pdf512.99 kBAdobe PDFDownload
Show simple item record

Google ScholarTM

Check


This item is available under the Attribution-NonCommercial-NoDerivs 3.0 Ireland. No item may be reproduced for commercial purposes. For other possible restrictions on use please refer to the publisher's URL where this is made available, or to notes contained in the item itself. Other terms may apply.