Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952425013> ?p ?o ?g. }
- W2952425013 abstract "Few-shot learning is a challenging problem where the goal is to achieve generalization from only few examples. Model-agnostic meta-learning (MAML) tackles the problem by formulating prior knowledge as a common initialization across tasks, which is then used to quickly adapt to unseen tasks. However, forcibly sharing an initialization can lead to conflicts among tasks and the compromised (undesired by tasks) location on optimization landscape, thereby hindering the task adaptation. Further, we observe that the degree of conflict differs among not only tasks but also layers of a neural network. Thus, we propose task-and-layer-wise attenuation on the compromised initialization to reduce its influence. As the attenuation dynamically controls (or selectively forgets) the influence of prior knowledge for a given task and each layer, we name our method as L2F (Learn to Forget). The experimental results demonstrate that the proposed method provides faster adaptation and greatly improves the performance. Furthermore, L2F can be easily applied and improve other state-of-the-art MAML-based frameworks, illustrating its simplicity and generalizability." @default.
- W2952425013 created "2019-06-27" @default.
- W2952425013 creator A5046504049 @default.
- W2952425013 creator A5048206537 @default.
- W2952425013 creator A5076005445 @default.
- W2952425013 date "2019-06-13" @default.
- W2952425013 modified "2023-09-27" @default.
- W2952425013 title "Learning to Forget for Meta-Learning" @default.
- W2952425013 cites W1598377843 @default.
- W2952425013 cites W1836465849 @default.
- W2952425013 cites W1849277567 @default.
- W2952425013 cites W2062118960 @default.
- W2952425013 cites W2089217417 @default.
- W2952425013 cites W2137825550 @default.
- W2952425013 cites W2158782408 @default.
- W2952425013 cites W2472819217 @default.
- W2952425013 cites W2601450892 @default.
- W2952425013 cites W2604117713 @default.
- W2952425013 cites W2742093937 @default.
- W2952425013 cites W2753160622 @default.
- W2952425013 cites W2787501667 @default.
- W2952425013 cites W2803672301 @default.
- W2952425013 cites W2892122929 @default.
- W2952425013 cites W2895671740 @default.
- W2952425013 cites W2904167876 @default.
- W2952425013 cites W2911248322 @default.
- W2952425013 cites W2912811302 @default.
- W2952425013 cites W2919053476 @default.
- W2952425013 cites W2949667497 @default.
- W2952425013 cites W2951775809 @default.
- W2952425013 cites W2962723986 @default.
- W2952425013 cites W2962835968 @default.
- W2952425013 cites W2962933129 @default.
- W2952425013 cites W2963303956 @default.
- W2952425013 cites W2963341924 @default.
- W2952425013 cites W2963399222 @default.
- W2952425013 cites W2963547174 @default.
- W2952425013 cites W2963551321 @default.
- W2952425013 cites W2963580001 @default.
- W2952425013 cites W2963641140 @default.
- W2952425013 cites W2963739978 @default.
- W2952425013 cites W2963959597 @default.
- W2952425013 cites W2964105864 @default.
- W2952425013 cites W2964112702 @default.
- W2952425013 cites W2964206659 @default.
- W2952425013 cites W3037856073 @default.
- W2952425013 cites W3091905774 @default.
- W2952425013 cites W3093329015 @default.
- W2952425013 cites W3118608800 @default.
- W2952425013 cites W99485931 @default.
- W2952425013 hasPublicationYear "2019" @default.
- W2952425013 type Work @default.
- W2952425013 sameAs 2952425013 @default.
- W2952425013 citedByCount "1" @default.
- W2952425013 countsByYear W29524250132020 @default.
- W2952425013 crossrefType "posted-content" @default.
- W2952425013 hasAuthorship W2952425013A5046504049 @default.
- W2952425013 hasAuthorship W2952425013A5048206537 @default.
- W2952425013 hasAuthorship W2952425013A5076005445 @default.
- W2952425013 hasConcept C105795698 @default.
- W2952425013 hasConcept C114466953 @default.
- W2952425013 hasConcept C119857082 @default.
- W2952425013 hasConcept C120665830 @default.
- W2952425013 hasConcept C121332964 @default.
- W2952425013 hasConcept C134306372 @default.
- W2952425013 hasConcept C139807058 @default.
- W2952425013 hasConcept C154945302 @default.
- W2952425013 hasConcept C162324750 @default.
- W2952425013 hasConcept C177148314 @default.
- W2952425013 hasConcept C178790620 @default.
- W2952425013 hasConcept C185592680 @default.
- W2952425013 hasConcept C187736073 @default.
- W2952425013 hasConcept C199360897 @default.
- W2952425013 hasConcept C27158222 @default.
- W2952425013 hasConcept C2779227376 @default.
- W2952425013 hasConcept C2780451532 @default.
- W2952425013 hasConcept C2781002164 @default.
- W2952425013 hasConcept C33923547 @default.
- W2952425013 hasConcept C41008148 @default.
- W2952425013 hasConcept C50644808 @default.
- W2952425013 hasConceptScore W2952425013C105795698 @default.
- W2952425013 hasConceptScore W2952425013C114466953 @default.
- W2952425013 hasConceptScore W2952425013C119857082 @default.
- W2952425013 hasConceptScore W2952425013C120665830 @default.
- W2952425013 hasConceptScore W2952425013C121332964 @default.
- W2952425013 hasConceptScore W2952425013C134306372 @default.
- W2952425013 hasConceptScore W2952425013C139807058 @default.
- W2952425013 hasConceptScore W2952425013C154945302 @default.
- W2952425013 hasConceptScore W2952425013C162324750 @default.
- W2952425013 hasConceptScore W2952425013C177148314 @default.
- W2952425013 hasConceptScore W2952425013C178790620 @default.
- W2952425013 hasConceptScore W2952425013C185592680 @default.
- W2952425013 hasConceptScore W2952425013C187736073 @default.
- W2952425013 hasConceptScore W2952425013C199360897 @default.
- W2952425013 hasConceptScore W2952425013C27158222 @default.
- W2952425013 hasConceptScore W2952425013C2779227376 @default.
- W2952425013 hasConceptScore W2952425013C2780451532 @default.
- W2952425013 hasConceptScore W2952425013C2781002164 @default.
- W2952425013 hasConceptScore W2952425013C33923547 @default.
- W2952425013 hasConceptScore W2952425013C41008148 @default.