Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287665610> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W4287665610 abstract "In real-world applications, data do not reflect the ones commonly used for neural networks training, since they are usually few, unlabeled and can be available as a stream. Hence many existing deep learning solutions suffer from a limited range of applications, in particular in the case of online streaming data that evolve over time. To narrow this gap, in this work we introduce a novel and complex setting involving unsupervised meta-continual learning with unbalanced tasks. These tasks are built through a clustering procedure applied to a fitted embedding space. We exploit a meta-learning scheme that simultaneously alleviates catastrophic forgetting and favors the generalization to new tasks. Moreover, to encourage feature reuse during the meta-optimization, we exploit a single inner loop taking advantage of an aggregated representation achieved through the use of a self-attention mechanism. Experimental results on few-shot learning benchmarks show competitive performance even compared to the supervised case. Additionally, we empirically observe that in an unsupervised scenario, the small tasks and the variability in the clusters pooling play a crucial role in the generalization capability of the network. Further, on complex datasets, the exploitation of more clusters than the true number of classes leads to higher results, even compared to the ones obtained with full supervision, suggesting that a predefined partitioning into classes can miss relevant structural information." @default.
- W4287665610 created "2022-07-25" @default.
- W4287665610 creator A5001856744 @default.
- W4287665610 creator A5058126024 @default.
- W4287665610 creator A5066187890 @default.
- W4287665610 creator A5075481810 @default.
- W4287665610 date "2020-09-17" @default.
- W4287665610 modified "2023-10-18" @default.
- W4287665610 title "Few-Shot Unsupervised Continual Learning through Meta-Examples" @default.
- W4287665610 doi "https://doi.org/10.48550/arxiv.2009.08107" @default.
- W4287665610 hasPublicationYear "2020" @default.
- W4287665610 type Work @default.
- W4287665610 citedByCount "0" @default.
- W4287665610 crossrefType "posted-content" @default.
- W4287665610 hasAuthorship W4287665610A5001856744 @default.
- W4287665610 hasAuthorship W4287665610A5058126024 @default.
- W4287665610 hasAuthorship W4287665610A5066187890 @default.
- W4287665610 hasAuthorship W4287665610A5075481810 @default.
- W4287665610 hasBestOaLocation W42876656101 @default.
- W4287665610 hasConcept C119857082 @default.
- W4287665610 hasConcept C120822770 @default.
- W4287665610 hasConcept C134306372 @default.
- W4287665610 hasConcept C138885662 @default.
- W4287665610 hasConcept C154945302 @default.
- W4287665610 hasConcept C162324750 @default.
- W4287665610 hasConcept C165696696 @default.
- W4287665610 hasConcept C177148314 @default.
- W4287665610 hasConcept C17744445 @default.
- W4287665610 hasConcept C187736073 @default.
- W4287665610 hasConcept C199539241 @default.
- W4287665610 hasConcept C2776359362 @default.
- W4287665610 hasConcept C2780451532 @default.
- W4287665610 hasConcept C2781002164 @default.
- W4287665610 hasConcept C33923547 @default.
- W4287665610 hasConcept C38652104 @default.
- W4287665610 hasConcept C41008148 @default.
- W4287665610 hasConcept C41608201 @default.
- W4287665610 hasConcept C41895202 @default.
- W4287665610 hasConcept C59404180 @default.
- W4287665610 hasConcept C70437156 @default.
- W4287665610 hasConcept C7149132 @default.
- W4287665610 hasConcept C73555534 @default.
- W4287665610 hasConcept C8038995 @default.
- W4287665610 hasConcept C94625758 @default.
- W4287665610 hasConceptScore W4287665610C119857082 @default.
- W4287665610 hasConceptScore W4287665610C120822770 @default.
- W4287665610 hasConceptScore W4287665610C134306372 @default.
- W4287665610 hasConceptScore W4287665610C138885662 @default.
- W4287665610 hasConceptScore W4287665610C154945302 @default.
- W4287665610 hasConceptScore W4287665610C162324750 @default.
- W4287665610 hasConceptScore W4287665610C165696696 @default.
- W4287665610 hasConceptScore W4287665610C177148314 @default.
- W4287665610 hasConceptScore W4287665610C17744445 @default.
- W4287665610 hasConceptScore W4287665610C187736073 @default.
- W4287665610 hasConceptScore W4287665610C199539241 @default.
- W4287665610 hasConceptScore W4287665610C2776359362 @default.
- W4287665610 hasConceptScore W4287665610C2780451532 @default.
- W4287665610 hasConceptScore W4287665610C2781002164 @default.
- W4287665610 hasConceptScore W4287665610C33923547 @default.
- W4287665610 hasConceptScore W4287665610C38652104 @default.
- W4287665610 hasConceptScore W4287665610C41008148 @default.
- W4287665610 hasConceptScore W4287665610C41608201 @default.
- W4287665610 hasConceptScore W4287665610C41895202 @default.
- W4287665610 hasConceptScore W4287665610C59404180 @default.
- W4287665610 hasConceptScore W4287665610C70437156 @default.
- W4287665610 hasConceptScore W4287665610C7149132 @default.
- W4287665610 hasConceptScore W4287665610C73555534 @default.
- W4287665610 hasConceptScore W4287665610C8038995 @default.
- W4287665610 hasConceptScore W4287665610C94625758 @default.
- W4287665610 hasLocation W42876656101 @default.
- W4287665610 hasOpenAccess W4287665610 @default.
- W4287665610 hasPrimaryLocation W42876656101 @default.
- W4287665610 hasRelatedWork W1491343858 @default.
- W4287665610 hasRelatedWork W2895106137 @default.
- W4287665610 hasRelatedWork W2912933387 @default.
- W4287665610 hasRelatedWork W2929676662 @default.
- W4287665610 hasRelatedWork W2964032613 @default.
- W4287665610 hasRelatedWork W2997669297 @default.
- W4287665610 hasRelatedWork W3086978629 @default.
- W4287665610 hasRelatedWork W3200361725 @default.
- W4287665610 hasRelatedWork W4287665610 @default.
- W4287665610 hasRelatedWork W4291381784 @default.
- W4287665610 isParatext "false" @default.
- W4287665610 isRetracted "false" @default.
- W4287665610 workType "article" @default.