Towards Causal Replay for Knowledge Rehearsal in Continual Learning
cam.depositDate | 2023-01-16 | |
dc.contributor.author | Churamani, N | |
dc.contributor.author | Cheong, J | |
dc.contributor.author | Kalkan, S | |
dc.contributor.author | Gunes, H | |
dc.contributor.orcid | Churamani, Nikhil [0000-0001-5926-0091] | |
dc.date.accessioned | 2023-01-17T00:31:29Z | |
dc.date.available | 2023-01-17T00:31:29Z | |
dc.date.issued | 2023 | |
dc.date.updated | 2023-01-16T01:03:01Z | |
dc.description.abstract | Given the challenges associated with the real-world deployment of Machine Learning (ML) models, especially towards efficiently integrating novel information on the go, both Continual Learning (CL) and Causality have been proposed and investigated individually as potent solutions. Despite their complementary nature, the bridge between them is still largely unexplored. In this work, we focus on causality to improve the learning and knowledge preservation capabilities of CL models. In particular, positing Causal Replay for knowledge rehearsal, we discuss how CL-based models can benefit from causal interventions towards improving their ability to replay past knowledge in order to mitigate forgetting. | |
dc.description.sponsorship | EPSRC, Alan Turing Institute, Cambridge Commonwealth Trust | |
dc.identifier.doi | 10.17863/CAM.92850 | |
dc.identifier.eissn | 2640-3498 | |
dc.identifier.uri | https://www.repository.cam.ac.uk/handle/1810/345428 | |
dc.language.iso | eng | |
dc.publisher | PMLR | |
dc.publisher.department | Department of Computer Science and Technology | |
dc.publisher.url | https://proceedings.mlr.press/v208/ | |
dc.rights | Attribution 4.0 International | |
dc.rights.uri | https://creativecommons.org/licenses/by/4.0/ | |
dc.subject | Causality | |
dc.subject | Continual Learning | |
dc.subject | Pseudo-rehearsal | |
dc.subject | Rehearsal | |
dc.title | Towards Causal Replay for Knowledge Rehearsal in Continual Learning | |
dc.type | Conference Object | |
dcterms.dateAccepted | 2022-12-05 | |
prism.publicationName | Proceedings of Machine Learning Research | |
pubs.conference-finish-date | 2023-02-08 | |
pubs.conference-name | Continual Causality Bridge Program at the AAAI Conference on Artificial Intelligence 2023 | |
pubs.conference-start-date | 2023-02-07 | |
pubs.funder-project-id | Engineering and Physical Sciences Research Council (EP/R030782/1) | |
pubs.funder-project-id | Alan Turing Institute (ATIPO000004438) | |
pubs.licence-display-name | Apollo Repository Deposit Licence Agreement | |
pubs.licence-identifier | apollo-deposit-licence-2-1 | |
rioxxterms.version | AM |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- AAAI_ContinualCausality_CausalReplay.pdf
- Size:
- 192.04 KB
- Format:
- Adobe Portable Document Format
- Description:
- Accepted version
- Licence
- https://creativecommons.org/licenses/by/4.0/