Matches in SemOpenAlex for { <https://semopenalex.org/work/W3155127799> ?p ?o ?g. }
Showing items 1 to 99 of
99
with 100 items per page.
- W3155127799 abstract "Multi-task learning(MTL) is an open and challenging problem in various real-world applications. The typical way of conducting multi-task learning is establishing some global parameter sharing mechanism across all tasks or assigning each task an individual set of parameters with cross-connections between tasks. However, for most existing approaches, all tasks just thoroughly or proportionally share all the features without distinguishing the helpfulness of them. By that, some tasks would be intervened by the unhelpful features that are useful for other tasks, leading to undesired negative transfer between tasks. In this paper, we design a novel architecture named the Multiple-level Sparse Sharing Model (MSSM), which can learn features selectively and share knowledge across all tasks efficiently. MSSM first employs a field-level sparse connection module (FSCM) to enable much more expressive combinations of feature fields to be learned for generalization across tasks while still allowing for task-specific features to be customized for each task. Furthermore, a cell-level sparse sharing module (CSSM) can recognize the sharing pattern through a set of coding variables that selectively choose which cells to route for a given task. Extensive experimental results on several real-world datasets show that MSSM outperforms SOTA models significantly in terms of AUC and LogLoss metrics." @default.
- W3155127799 created "2021-04-26" @default.
- W3155127799 creator A5010085095 @default.
- W3155127799 creator A5012431501 @default.
- W3155127799 creator A5017443217 @default.
- W3155127799 creator A5030849407 @default.
- W3155127799 creator A5044628261 @default.
- W3155127799 creator A5064691231 @default.
- W3155127799 creator A5066599847 @default.
- W3155127799 creator A5069190300 @default.
- W3155127799 creator A5071951805 @default.
- W3155127799 creator A5082449442 @default.
- W3155127799 creator A5088720980 @default.
- W3155127799 date "2021-07-11" @default.
- W3155127799 modified "2023-10-18" @default.
- W3155127799 title "MSSM: A Multiple-level Sparse Sharing Model for Efficient Multi-Task Learning" @default.
- W3155127799 cites W2117130368 @default.
- W3155127799 cites W2809290718 @default.
- W3155127799 cites W2903852246 @default.
- W3155127799 cites W2963430933 @default.
- W3155127799 cites W2963877604 @default.
- W3155127799 cites W2964089273 @default.
- W3155127799 cites W3034704453 @default.
- W3155127799 cites W3035596828 @default.
- W3155127799 cites W3080189354 @default.
- W3155127799 cites W3087931390 @default.
- W3155127799 doi "https://doi.org/10.1145/3404835.3463022" @default.
- W3155127799 hasPublicationYear "2021" @default.
- W3155127799 type Work @default.
- W3155127799 sameAs 3155127799 @default.
- W3155127799 citedByCount "16" @default.
- W3155127799 countsByYear W31551277992021 @default.
- W3155127799 countsByYear W31551277992022 @default.
- W3155127799 countsByYear W31551277992023 @default.
- W3155127799 crossrefType "proceedings-article" @default.
- W3155127799 hasAuthorship W3155127799A5010085095 @default.
- W3155127799 hasAuthorship W3155127799A5012431501 @default.
- W3155127799 hasAuthorship W3155127799A5017443217 @default.
- W3155127799 hasAuthorship W3155127799A5030849407 @default.
- W3155127799 hasAuthorship W3155127799A5044628261 @default.
- W3155127799 hasAuthorship W3155127799A5064691231 @default.
- W3155127799 hasAuthorship W3155127799A5066599847 @default.
- W3155127799 hasAuthorship W3155127799A5069190300 @default.
- W3155127799 hasAuthorship W3155127799A5071951805 @default.
- W3155127799 hasAuthorship W3155127799A5082449442 @default.
- W3155127799 hasAuthorship W3155127799A5088720980 @default.
- W3155127799 hasConcept C119857082 @default.
- W3155127799 hasConcept C134306372 @default.
- W3155127799 hasConcept C138885662 @default.
- W3155127799 hasConcept C150899416 @default.
- W3155127799 hasConcept C154945302 @default.
- W3155127799 hasConcept C162324750 @default.
- W3155127799 hasConcept C177148314 @default.
- W3155127799 hasConcept C177264268 @default.
- W3155127799 hasConcept C187736073 @default.
- W3155127799 hasConcept C199360897 @default.
- W3155127799 hasConcept C202444582 @default.
- W3155127799 hasConcept C2776401178 @default.
- W3155127799 hasConcept C2780451532 @default.
- W3155127799 hasConcept C28006648 @default.
- W3155127799 hasConcept C33923547 @default.
- W3155127799 hasConcept C41008148 @default.
- W3155127799 hasConcept C41895202 @default.
- W3155127799 hasConcept C9652623 @default.
- W3155127799 hasConceptScore W3155127799C119857082 @default.
- W3155127799 hasConceptScore W3155127799C134306372 @default.
- W3155127799 hasConceptScore W3155127799C138885662 @default.
- W3155127799 hasConceptScore W3155127799C150899416 @default.
- W3155127799 hasConceptScore W3155127799C154945302 @default.
- W3155127799 hasConceptScore W3155127799C162324750 @default.
- W3155127799 hasConceptScore W3155127799C177148314 @default.
- W3155127799 hasConceptScore W3155127799C177264268 @default.
- W3155127799 hasConceptScore W3155127799C187736073 @default.
- W3155127799 hasConceptScore W3155127799C199360897 @default.
- W3155127799 hasConceptScore W3155127799C202444582 @default.
- W3155127799 hasConceptScore W3155127799C2776401178 @default.
- W3155127799 hasConceptScore W3155127799C2780451532 @default.
- W3155127799 hasConceptScore W3155127799C28006648 @default.
- W3155127799 hasConceptScore W3155127799C33923547 @default.
- W3155127799 hasConceptScore W3155127799C41008148 @default.
- W3155127799 hasConceptScore W3155127799C41895202 @default.
- W3155127799 hasConceptScore W3155127799C9652623 @default.
- W3155127799 hasLocation W31551277991 @default.
- W3155127799 hasOpenAccess W3155127799 @default.
- W3155127799 hasPrimaryLocation W31551277991 @default.
- W3155127799 hasRelatedWork W2946016983 @default.
- W3155127799 hasRelatedWork W2960456850 @default.
- W3155127799 hasRelatedWork W3010482781 @default.
- W3155127799 hasRelatedWork W3021430260 @default.
- W3155127799 hasRelatedWork W3090469165 @default.
- W3155127799 hasRelatedWork W4281645081 @default.
- W3155127799 hasRelatedWork W4308262314 @default.
- W3155127799 hasRelatedWork W4312200629 @default.
- W3155127799 hasRelatedWork W4318718973 @default.
- W3155127799 hasRelatedWork W4382286161 @default.
- W3155127799 isParatext "false" @default.
- W3155127799 isRetracted "false" @default.
- W3155127799 magId "3155127799" @default.
- W3155127799 workType "article" @default.