Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387348566> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4387348566 abstract "Artificial neural networks suffer from catastrophic forgetting when they are sequentially trained on multiple tasks. To overcome this problem, there exist many continual learning strategies. One of the most effective is the hypernetwork-based approach. The hypernetwork generates the weights of a target model based on the task's identity. The model's main limitation is that hypernetwork can produce completely different nests for each task. Consequently, each task is solved separately. The model does not use information from the network dedicated to previous tasks and practically produces new architectures when it learns the subsequent tasks. To solve such a problem, we use the lottery ticket hypothesis, which postulates the existence of sparse subnetworks, named winning tickets, that preserve the performance of a full network. In the paper, we propose a method called HyperMask, which trains a single network for all tasks. Hypernetwork produces semi-binary masks to obtain target subnetworks dedicated to new tasks. This solution inherits the ability of the hypernetwork to adapt to new tasks with minimal forgetting. Moreover, due to the lottery ticket hypothesis, we can use a single network with weighted subnets dedicated to each task." @default.
- W4387348566 created "2023-10-05" @default.
- W4387348566 creator A5034373932 @default.
- W4387348566 creator A5068511223 @default.
- W4387348566 date "2023-09-29" @default.
- W4387348566 modified "2023-10-13" @default.
- W4387348566 title "HyperMask: Adaptive Hypernetwork-based Masks for Continual Learning" @default.
- W4387348566 doi "https://doi.org/10.48550/arxiv.2310.00113" @default.
- W4387348566 hasPublicationYear "2023" @default.
- W4387348566 type Work @default.
- W4387348566 citedByCount "0" @default.
- W4387348566 crossrefType "posted-content" @default.
- W4387348566 hasAuthorship W4387348566A5034373932 @default.
- W4387348566 hasAuthorship W4387348566A5068511223 @default.
- W4387348566 hasBestOaLocation W43873485661 @default.
- W4387348566 hasConcept C119857082 @default.
- W4387348566 hasConcept C138885662 @default.
- W4387348566 hasConcept C154945302 @default.
- W4387348566 hasConcept C162324750 @default.
- W4387348566 hasConcept C187736073 @default.
- W4387348566 hasConcept C190839683 @default.
- W4387348566 hasConcept C205649164 @default.
- W4387348566 hasConcept C2780451532 @default.
- W4387348566 hasConcept C41008148 @default.
- W4387348566 hasConcept C41895202 @default.
- W4387348566 hasConcept C50644808 @default.
- W4387348566 hasConcept C58640448 @default.
- W4387348566 hasConcept C7149132 @default.
- W4387348566 hasConcept C80444323 @default.
- W4387348566 hasConceptScore W4387348566C119857082 @default.
- W4387348566 hasConceptScore W4387348566C138885662 @default.
- W4387348566 hasConceptScore W4387348566C154945302 @default.
- W4387348566 hasConceptScore W4387348566C162324750 @default.
- W4387348566 hasConceptScore W4387348566C187736073 @default.
- W4387348566 hasConceptScore W4387348566C190839683 @default.
- W4387348566 hasConceptScore W4387348566C205649164 @default.
- W4387348566 hasConceptScore W4387348566C2780451532 @default.
- W4387348566 hasConceptScore W4387348566C41008148 @default.
- W4387348566 hasConceptScore W4387348566C41895202 @default.
- W4387348566 hasConceptScore W4387348566C50644808 @default.
- W4387348566 hasConceptScore W4387348566C58640448 @default.
- W4387348566 hasConceptScore W4387348566C7149132 @default.
- W4387348566 hasConceptScore W4387348566C80444323 @default.
- W4387348566 hasLocation W43873485661 @default.
- W4387348566 hasOpenAccess W4387348566 @default.
- W4387348566 hasPrimaryLocation W43873485661 @default.
- W4387348566 hasRelatedWork W2104218666 @default.
- W4387348566 hasRelatedWork W2145559838 @default.
- W4387348566 hasRelatedWork W2164121020 @default.
- W4387348566 hasRelatedWork W2794885965 @default.
- W4387348566 hasRelatedWork W2959635497 @default.
- W4387348566 hasRelatedWork W3116498279 @default.
- W4387348566 hasRelatedWork W3183027292 @default.
- W4387348566 hasRelatedWork W4287549553 @default.
- W4387348566 hasRelatedWork W4289718052 @default.
- W4387348566 hasRelatedWork W4310285384 @default.
- W4387348566 isParatext "false" @default.
- W4387348566 isRetracted "false" @default.
- W4387348566 workType "article" @default.