Beyond Catastrophic Forgetting in Continual Learning: An Attempt with SVM

Simple item page
aut.relation.conferenceICMLen_NZ
aut.researcherBenavides Prado, Diana
dc.contributor.authorBenavides Prado, Den_NZ
dc.date.accessioned2020-11-26T22:43:00Z
dc.date.available2020-11-26T22:43:00Z
dc.date.copyright2020-07-18en_NZ
dc.date.issued2020-07-18en_NZ
dc.description.abstractA big challenge in continual learning is avoiding catastrophically forgetting previously learned tasks. The possibility of improving existing knowledge whilst integrating new information has been much less explored. In this paper we describe a method that aims to improve the performance of previously learned tasks by refining their knowledge while new tasks are observed. Our method is an example of this ability in the context of Support Vector Machines for binary classification tasks, which encourages retention of existing knowledge whilst refining. Experiments with synthetic and real-world datasets show that the performance of these tasks can be continually improved by transferring selected knowledge, leading to the improvement on the performance of the learning system as a whole.
dc.identifier.citationProceedings of the 37 th International Conference on Machine Learning, Vienna, Austria, PMLR 108, 2020.
dc.identifier.urihttps://hdl.handle.net/10292/13829
dc.publisherThe International Conference on Machine Learning (ICML)
dc.relation.urihttps://icml.cc/Conferences/2020/ScheduleMultitrack?event=5743
dc.rightsCopyright 2020 by the author(s).
dc.rights.accessrightsOpenAccessen_NZ
dc.titleBeyond Catastrophic Forgetting in Continual Learning: An Attempt with SVMen_NZ
dc.typeConference Contribution
pubs.elements-id393535
pubs.organisational-data/AUT
pubs.organisational-data/AUT/Culture & Society
pubs.organisational-data/AUT/Culture & Society/Social Science & Public Policy
pubs.organisational-data/AUT/PBRF
pubs.organisational-data/AUT/PBRF/PBRF Business Economics and Law
pubs.organisational-data/AUT/PBRF/PBRF Business Economics and Law/Faculty Review Team PBRF 2018
pubs.organisational-data/AUT/PBRF/PBRF Business Economics and Law/School of Economics PBRF 2018
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
9_CameraReadySubmission_ICMLCL (1).pdf
Size:
364.74 KB
Format:
Adobe Portable Document Format
Description:
Conference Contribution
Download
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
AUT Grant of Licence for Tuwhera Aug 2018.pdf
Size:
276.29 KB
Format:
Adobe Portable Document Format
Description:
Download
Collections
School of Social Sciences and Public Policy