FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2971296719> ?p ?o ?g. }

• W2971296719 endingPage "5857" @default.
• W2971296719 startingPage "5846" @default.
• W2971296719 abstract "In multi-task learning, a major challenge springs from a notorious issue known as negative transfer, which refers to the phenomenon that sharing the knowledge with dissimilar and hard tasks often results in a worsened performance. To circumvent this issue, we propose a novel multi-task learning method, which simultaneously learns latent task representations and a block-diagonal Latent Task Assignment Matrix (LTAM). Different from most of the previous work, pursuing the Block-Diagonal structure of LTAM (assigning latent tasks to output tasks) alleviates negative transfer via collaboratively grouping latent tasks and output tasks such that inter-group knowledge transfer and sharing is suppressed. This goal is challenging, since 1) our notion of Block-Diagonal Property extends the traditional notion for square matrices where the $i$-th column and the $i$-th column represents the same concept; 2) marginal constraints on rows and columns are also required for avoiding isolated latent/output tasks. Facing such challenges, we propose a novel regularizer by means of an equivalent spectral condition realizing this generalized block-diagonal property. Practically, we provide a relaxation scheme which improves the flexibility of the model. With the objective function given, we then propose an alternating optimization method, which not only tells how negative transfer is alleviated in our method but also reveals an interesting connection between our method and the optimal transport problem. Finally, the method is demonstrated on a simulation dataset, three real-world benchmark datasets and further applied to personalized attribute predictions." @default.
• W2971296719 created "2019-09-05" @default.
• W2971296719 creator A5009753955 @default.
• W2971296719 creator A5028597017 @default.
• W2971296719 creator A5053213796 @default.
• W2971296719 creator A5068837264 @default.
• W2971296719 creator A5089092566 @default.
• W2971296719 date "2019-01-01" @default.
• W2971296719 modified "2023-09-24" @default.
• W2971296719 title "Generalized Block-Diagonal Structure Pursuit: Learning Soft Latent Task Assignment against Negative Transfer" @default.
• W2971296719 cites W1585468632 @default.
• W2971296719 cites W1647603301 @default.
• W2971296719 cites W1939941161 @default.
• W2971296719 cites W1963753244 @default.
• W2971296719 cites W1995766682 @default.
• W2971296719 cites W2005126631 @default.
• W2971296719 cites W2015492922 @default.
• W2971296719 cites W2033365921 @default.
• W2971296719 cites W2065180801 @default.
• W2971296719 cites W2110742393 @default.
• W2971296719 cites W2118099552 @default.
• W2971296719 cites W2150287743 @default.
• W2971296719 cites W2170563643 @default.
• W2971296719 cites W2183341477 @default.
• W2971296719 cites W2186054958 @default.
• W2971296719 cites W2187089797 @default.
• W2971296719 cites W2561587488 @default.
• W2971296719 cites W2615692152 @default.
• W2971296719 cites W2741887298 @default.
• W2971296719 cites W2793477525 @default.
• W2971296719 cites W2798441115 @default.
• W2971296719 cites W2798685991 @default.
• W2971296719 cites W2798778171 @default.
• W2971296719 cites W2809497023 @default.
• W2971296719 cites W2809623940 @default.
• W2971296719 cites W2891595829 @default.
• W2971296719 cites W2897874959 @default.
• W2971296719 cites W2904759072 @default.
• W2971296719 cites W2913340405 @default.
• W2971296719 cites W2943996654 @default.
• W2971296719 cites W2949201716 @default.
• W2971296719 cites W2950094835 @default.
• W2971296719 cites W2953126679 @default.
• W2971296719 cites W2962736834 @default.
• W2971296719 cites W2963424430 @default.
• W2971296719 cites W2963449770 @default.
• W2971296719 cites W2964252913 @default.
• W2971296719 cites W2981663715 @default.
• W2971296719 cites W344512287 @default.
• W2971296719 cites W2962764337 @default.
• W2971296719 hasPublicationYear "2019" @default.
• W2971296719 type Work @default.
• W2971296719 sameAs 2971296719 @default.
• W2971296719 citedByCount "0" @default.
• W2971296719 crossrefType "proceedings-article" @default.
• W2971296719 hasAuthorship W2971296719A5009753955 @default.
• W2971296719 hasAuthorship W2971296719A5028597017 @default.
• W2971296719 hasAuthorship W2971296719A5053213796 @default.
• W2971296719 hasAuthorship W2971296719A5068837264 @default.
• W2971296719 hasAuthorship W2971296719A5089092566 @default.
• W2971296719 hasConcept C105795698 @default.
• W2971296719 hasConcept C106487976 @default.
• W2971296719 hasConcept C11413529 @default.
• W2971296719 hasConcept C121332964 @default.
• W2971296719 hasConcept C130367717 @default.
• W2971296719 hasConcept C13280743 @default.
• W2971296719 hasConcept C150899416 @default.
• W2971296719 hasConcept C154945302 @default.
• W2971296719 hasConcept C15744967 @default.
• W2971296719 hasConcept C158693339 @default.
• W2971296719 hasConcept C159985019 @default.
• W2971296719 hasConcept C162324750 @default.
• W2971296719 hasConcept C185798385 @default.
• W2971296719 hasConcept C187736073 @default.
• W2971296719 hasConcept C192562407 @default.
• W2971296719 hasConcept C205649164 @default.
• W2971296719 hasConcept C2524010 @default.
• W2971296719 hasConcept C2776029896 @default.
• W2971296719 hasConcept C2777210771 @default.
• W2971296719 hasConcept C2780451532 @default.
• W2971296719 hasConcept C2780598303 @default.
• W2971296719 hasConcept C28006648 @default.
• W2971296719 hasConcept C33923547 @default.
• W2971296719 hasConcept C41008148 @default.
• W2971296719 hasConcept C51167844 @default.
• W2971296719 hasConcept C62520636 @default.
• W2971296719 hasConcept C77805123 @default.
• W2971296719 hasConcept C85817219 @default.
• W2971296719 hasConceptScore W2971296719C105795698 @default.
• W2971296719 hasConceptScore W2971296719C106487976 @default.
• W2971296719 hasConceptScore W2971296719C11413529 @default.
• W2971296719 hasConceptScore W2971296719C121332964 @default.
• W2971296719 hasConceptScore W2971296719C130367717 @default.
• W2971296719 hasConceptScore W2971296719C13280743 @default.
• W2971296719 hasConceptScore W2971296719C150899416 @default.
• W2971296719 hasConceptScore W2971296719C154945302 @default.
• W2971296719 hasConceptScore W2971296719C15744967 @default.
• W2971296719 hasConceptScore W2971296719C158693339 @default.

• next