FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3204408769> ?p ?o ?g. }

• W3204408769 endingPage "107535" @default.
• W3204408769 startingPage "107535" @default.
• W3204408769 abstract "Deep reinforcement learning (DRL) has achieved great success in recent years by combining the feature extraction power of deep learning and the decision power of reinforcement learning techniques. In the literature, Convolutional Neural Networks (CNN) is usually used as the feature extraction method and recent studies have shown that the performances of the DRL algorithms can be greatly improved with the utilization of the attention mechanism, where the raw attentions are directly used for the decision-making. However, as is well-known, reinforcement learning is a trial-and-error process and it is almost impossible to learn an optimal policy in the beginning of the learning, especially in environments with sparse rewards, which in turn will cause the raw attention-based models can only remember and utilize the attention information indiscriminately for different areas and may focus on some task-irrelevant regions, but the focusing on such task-irrelevant regions is usually helpless and ineffective for the agent to find the optimal policy. To address this issue, we propose a gated multi-attention mechanism, which is then combined with the Deep Q-learning network (GMAQN). The gated multi-attention representation module (GMA) in GMAQN can effectively eliminate task-irrelevant attention information in the early phase of the trial-and-error process and improve the stability of the model. The proposed method has been demonstrated on the challenging domain of classic Atari 2600 games and experimental results show that compared with the baselines, our method can achieve better performance in terms of both the scores and the effect of focusing in the key regions." @default.
• W3204408769 created "2021-10-11" @default.
• W3204408769 creator A5019574542 @default.
• W3204408769 creator A5051623896 @default.
• W3204408769 creator A5053996029 @default.
• W3204408769 date "2021-12-01" @default.
• W3204408769 modified "2023-10-16" @default.
• W3204408769 title "Gated multi-attention representation in reinforcement learning" @default.
• W3204408769 cites W1689711448 @default.
• W3204408769 cites W2064675550 @default.
• W3204408769 cites W2115293355 @default.
• W3204408769 cites W2119259385 @default.
• W3204408769 cites W2145339207 @default.
• W3204408769 cites W2761873684 @default.
• W3204408769 cites W2766447205 @default.
• W3204408769 cites W2776507577 @default.
• W3204408769 cites W2787259794 @default.
• W3204408769 cites W2904764169 @default.
• W3204408769 cites W2952472710 @default.
• W3204408769 cites W2962858109 @default.
• W3204408769 cites W2962949934 @default.
• W3204408769 cites W3008901316 @default.
• W3204408769 cites W3036066034 @default.
• W3204408769 cites W3101438731 @default.
• W3204408769 cites W3106649810 @default.
• W3204408769 cites W3124951096 @default.
• W3204408769 cites W3175752186 @default.
• W3204408769 cites W32403112 @default.
• W3204408769 cites W4253365321 @default.
• W3204408769 cites W4362203700 @default.
• W3204408769 doi "https://doi.org/10.1016/j.knosys.2021.107535" @default.
• W3204408769 hasPublicationYear "2021" @default.
• W3204408769 type Work @default.
• W3204408769 sameAs 3204408769 @default.
• W3204408769 citedByCount "4" @default.
• W3204408769 countsByYear W32044087692021 @default.
• W3204408769 countsByYear W32044087692022 @default.
• W3204408769 crossrefType "journal-article" @default.
• W3204408769 hasAuthorship W3204408769A5019574542 @default.
• W3204408769 hasAuthorship W3204408769A5051623896 @default.
• W3204408769 hasAuthorship W3204408769A5053996029 @default.
• W3204408769 hasConcept C108583219 @default.
• W3204408769 hasConcept C111919701 @default.
• W3204408769 hasConcept C112972136 @default.
• W3204408769 hasConcept C119857082 @default.
• W3204408769 hasConcept C120665830 @default.
• W3204408769 hasConcept C121332964 @default.
• W3204408769 hasConcept C138885662 @default.
• W3204408769 hasConcept C154945302 @default.
• W3204408769 hasConcept C162324750 @default.
• W3204408769 hasConcept C17744445 @default.
• W3204408769 hasConcept C187736073 @default.
• W3204408769 hasConcept C192209626 @default.
• W3204408769 hasConcept C199539241 @default.
• W3204408769 hasConcept C2776359362 @default.
• W3204408769 hasConcept C2776401178 @default.
• W3204408769 hasConcept C2780451532 @default.
• W3204408769 hasConcept C28006648 @default.
• W3204408769 hasConcept C41008148 @default.
• W3204408769 hasConcept C41895202 @default.
• W3204408769 hasConcept C59404180 @default.
• W3204408769 hasConcept C81363708 @default.
• W3204408769 hasConcept C94625758 @default.
• W3204408769 hasConcept C97541855 @default.
• W3204408769 hasConcept C98045186 @default.
• W3204408769 hasConceptScore W3204408769C108583219 @default.
• W3204408769 hasConceptScore W3204408769C111919701 @default.
• W3204408769 hasConceptScore W3204408769C112972136 @default.
• W3204408769 hasConceptScore W3204408769C119857082 @default.
• W3204408769 hasConceptScore W3204408769C120665830 @default.
• W3204408769 hasConceptScore W3204408769C121332964 @default.
• W3204408769 hasConceptScore W3204408769C138885662 @default.
• W3204408769 hasConceptScore W3204408769C154945302 @default.
• W3204408769 hasConceptScore W3204408769C162324750 @default.
• W3204408769 hasConceptScore W3204408769C17744445 @default.
• W3204408769 hasConceptScore W3204408769C187736073 @default.
• W3204408769 hasConceptScore W3204408769C192209626 @default.
• W3204408769 hasConceptScore W3204408769C199539241 @default.
• W3204408769 hasConceptScore W3204408769C2776359362 @default.
• W3204408769 hasConceptScore W3204408769C2776401178 @default.
• W3204408769 hasConceptScore W3204408769C2780451532 @default.
• W3204408769 hasConceptScore W3204408769C28006648 @default.
• W3204408769 hasConceptScore W3204408769C41008148 @default.
• W3204408769 hasConceptScore W3204408769C41895202 @default.
• W3204408769 hasConceptScore W3204408769C59404180 @default.
• W3204408769 hasConceptScore W3204408769C81363708 @default.
• W3204408769 hasConceptScore W3204408769C94625758 @default.
• W3204408769 hasConceptScore W3204408769C97541855 @default.
• W3204408769 hasConceptScore W3204408769C98045186 @default.
• W3204408769 hasFunder F4320321001 @default.
• W3204408769 hasFunder F4320322725 @default.
• W3204408769 hasLocation W32044087691 @default.
• W3204408769 hasOpenAccess W3204408769 @default.
• W3204408769 hasPrimaryLocation W32044087691 @default.
• W3204408769 hasRelatedWork W2337926734 @default.
• W3204408769 hasRelatedWork W2597787948 @default.
• W3204408769 hasRelatedWork W2768413403 @default.
• W3204408769 hasRelatedWork W2908875379 @default.

• next