Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386907402> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4386907402 abstract "Recommender systems usually leverage multi-task learning methods to simultaneously optimize several objectives because of the multi-faceted user behavior data. The typical way of conducting multi-task learning is to establish appropriate parameter sharing across multiple tasks at lower layers while reserving a separate task tower for each task at upper layers. Since the task towers exert direct impact on the prediction results, we argue that the architecture of standalone task towers is sub-optimal for promoting positive knowledge sharing. Accordingly, we propose the framework of Deep Mutual Learning across task towers, which is compatible with various backbone multi-task networks. Extensive offline experiments and online AB tests are conducted to evaluate and verify the proposed approach's effectiveness." @default.
- W4386907402 created "2023-09-21" @default.
- W4386907402 creator A5001101977 @default.
- W4386907402 creator A5007690810 @default.
- W4386907402 creator A5021827773 @default.
- W4386907402 creator A5070388252 @default.
- W4386907402 date "2023-09-19" @default.
- W4386907402 modified "2023-10-16" @default.
- W4386907402 title "Deep Mutual Learning across Task Towers for Effective Multi-Task Recommender Learning" @default.
- W4386907402 doi "https://doi.org/10.48550/arxiv.2309.10357" @default.
- W4386907402 hasPublicationYear "2023" @default.
- W4386907402 type Work @default.
- W4386907402 citedByCount "0" @default.
- W4386907402 crossrefType "posted-content" @default.
- W4386907402 hasAuthorship W4386907402A5001101977 @default.
- W4386907402 hasAuthorship W4386907402A5007690810 @default.
- W4386907402 hasAuthorship W4386907402A5021827773 @default.
- W4386907402 hasAuthorship W4386907402A5070388252 @default.
- W4386907402 hasBestOaLocation W43869074021 @default.
- W4386907402 hasConcept C108583219 @default.
- W4386907402 hasConcept C119857082 @default.
- W4386907402 hasConcept C127413603 @default.
- W4386907402 hasConcept C153083717 @default.
- W4386907402 hasConcept C154945302 @default.
- W4386907402 hasConcept C201995342 @default.
- W4386907402 hasConcept C2776604539 @default.
- W4386907402 hasConcept C2780451532 @default.
- W4386907402 hasConcept C28006648 @default.
- W4386907402 hasConcept C41008148 @default.
- W4386907402 hasConcept C557471498 @default.
- W4386907402 hasConcept C56739046 @default.
- W4386907402 hasConceptScore W4386907402C108583219 @default.
- W4386907402 hasConceptScore W4386907402C119857082 @default.
- W4386907402 hasConceptScore W4386907402C127413603 @default.
- W4386907402 hasConceptScore W4386907402C153083717 @default.
- W4386907402 hasConceptScore W4386907402C154945302 @default.
- W4386907402 hasConceptScore W4386907402C201995342 @default.
- W4386907402 hasConceptScore W4386907402C2776604539 @default.
- W4386907402 hasConceptScore W4386907402C2780451532 @default.
- W4386907402 hasConceptScore W4386907402C28006648 @default.
- W4386907402 hasConceptScore W4386907402C41008148 @default.
- W4386907402 hasConceptScore W4386907402C557471498 @default.
- W4386907402 hasConceptScore W4386907402C56739046 @default.
- W4386907402 hasLocation W43869074021 @default.
- W4386907402 hasOpenAccess W4386907402 @default.
- W4386907402 hasPrimaryLocation W43869074021 @default.
- W4386907402 hasRelatedWork W2597787948 @default.
- W4386907402 hasRelatedWork W3047894882 @default.
- W4386907402 hasRelatedWork W3200098538 @default.
- W4386907402 hasRelatedWork W3208584567 @default.
- W4386907402 hasRelatedWork W4221031031 @default.
- W4386907402 hasRelatedWork W4223943233 @default.
- W4386907402 hasRelatedWork W4295854770 @default.
- W4386907402 hasRelatedWork W4312200629 @default.
- W4386907402 hasRelatedWork W4360585206 @default.
- W4386907402 hasRelatedWork W4380075502 @default.
- W4386907402 isParatext "false" @default.
- W4386907402 isRetracted "false" @default.
- W4386907402 workType "article" @default.