Matches in SemOpenAlex for { <https://semopenalex.org/work/W3042765459> ?p ?o ?g. }
- W3042765459 abstract "Experience replay, which enables the agents to remember and reuse experience from the past, plays a significant role in the success of off-policy reinforcement learning (RL). To utilize the experience replay efficiently, experience transitions should be sampled with consideration of their significance, such that the known prioritized experience replay (PER) further allows to sample more important experience. Yet, the conventional PER may result in generating highly biased samples due to considering a single metric such as TD-error and computing the sampling rate independently for each experience. To tackle this issue, we propose a Neural Experience Replay Sampler (NERS), which adaptively evaluates the relative importance of a sampled transition by obtaining context from not only its (local) values that characterize itself such as TD-error or the raw features but also other (global) transitions. We validate our framework on multiple benchmark tasks for both continuous and discrete controls and show that the proposed framework significantly improves the performance of various off-policy RL methods. Further analysis confirms that the improvements indeed come from the use of diverse features and the consideration of the relative importance of experiences." @default.
- W3042765459 created "2020-07-23" @default.
- W3042765459 creator A5003384121 @default.
- W3042765459 creator A5033571780 @default.
- W3042765459 creator A5035397475 @default.
- W3042765459 creator A5070302452 @default.
- W3042765459 creator A5086698569 @default.
- W3042765459 date "2020-07-14" @default.
- W3042765459 modified "2023-10-07" @default.
- W3042765459 title "Learning to Sample with Local and Global Contexts in Experience Replay Buffer" @default.
- W3042765459 cites W1522301498 @default.
- W3042765459 cites W2114384389 @default.
- W3042765459 cites W2119717200 @default.
- W3042765459 cites W2145339207 @default.
- W3042765459 cites W2158782408 @default.
- W3042765459 cites W2165150801 @default.
- W3042765459 cites W2551887912 @default.
- W3042765459 cites W2772721022 @default.
- W3042765459 cites W2775482448 @default.
- W3042765459 cites W2904246096 @default.
- W3042765459 cites W2905810301 @default.
- W3042765459 cites W2946901134 @default.
- W3042765459 cites W2948708918 @default.
- W3042765459 cites W2951799221 @default.
- W3042765459 cites W2962902376 @default.
- W3042765459 cites W2963477884 @default.
- W3042765459 cites W2963733916 @default.
- W3042765459 cites W2963864421 @default.
- W3042765459 cites W2963923407 @default.
- W3042765459 cites W2964048876 @default.
- W3042765459 cites W2964291307 @default.
- W3042765459 cites W2965870268 @default.
- W3042765459 cites W2966234803 @default.
- W3042765459 cites W2969281871 @default.
- W3042765459 cites W2970202659 @default.
- W3042765459 cites W2970586779 @default.
- W3042765459 cites W2971204130 @default.
- W3042765459 cites W3005871422 @default.
- W3042765459 cites W3034971196 @default.
- W3042765459 cites W3089091950 @default.
- W3042765459 hasPublicationYear "2020" @default.
- W3042765459 type Work @default.
- W3042765459 sameAs 3042765459 @default.
- W3042765459 citedByCount "0" @default.
- W3042765459 crossrefType "posted-content" @default.
- W3042765459 hasAuthorship W3042765459A5003384121 @default.
- W3042765459 hasAuthorship W3042765459A5033571780 @default.
- W3042765459 hasAuthorship W3042765459A5035397475 @default.
- W3042765459 hasAuthorship W3042765459A5070302452 @default.
- W3042765459 hasAuthorship W3042765459A5086698569 @default.
- W3042765459 hasConcept C119857082 @default.
- W3042765459 hasConcept C13280743 @default.
- W3042765459 hasConcept C140779682 @default.
- W3042765459 hasConcept C151730666 @default.
- W3042765459 hasConcept C154945302 @default.
- W3042765459 hasConcept C162324750 @default.
- W3042765459 hasConcept C176217482 @default.
- W3042765459 hasConcept C185592680 @default.
- W3042765459 hasConcept C185798385 @default.
- W3042765459 hasConcept C18903297 @default.
- W3042765459 hasConcept C198531522 @default.
- W3042765459 hasConcept C205649164 @default.
- W3042765459 hasConcept C206588197 @default.
- W3042765459 hasConcept C21547014 @default.
- W3042765459 hasConcept C2779343474 @default.
- W3042765459 hasConcept C41008148 @default.
- W3042765459 hasConcept C43617362 @default.
- W3042765459 hasConcept C76155785 @default.
- W3042765459 hasConcept C86803240 @default.
- W3042765459 hasConcept C94915269 @default.
- W3042765459 hasConcept C97541855 @default.
- W3042765459 hasConceptScore W3042765459C119857082 @default.
- W3042765459 hasConceptScore W3042765459C13280743 @default.
- W3042765459 hasConceptScore W3042765459C140779682 @default.
- W3042765459 hasConceptScore W3042765459C151730666 @default.
- W3042765459 hasConceptScore W3042765459C154945302 @default.
- W3042765459 hasConceptScore W3042765459C162324750 @default.
- W3042765459 hasConceptScore W3042765459C176217482 @default.
- W3042765459 hasConceptScore W3042765459C185592680 @default.
- W3042765459 hasConceptScore W3042765459C185798385 @default.
- W3042765459 hasConceptScore W3042765459C18903297 @default.
- W3042765459 hasConceptScore W3042765459C198531522 @default.
- W3042765459 hasConceptScore W3042765459C205649164 @default.
- W3042765459 hasConceptScore W3042765459C206588197 @default.
- W3042765459 hasConceptScore W3042765459C21547014 @default.
- W3042765459 hasConceptScore W3042765459C2779343474 @default.
- W3042765459 hasConceptScore W3042765459C41008148 @default.
- W3042765459 hasConceptScore W3042765459C43617362 @default.
- W3042765459 hasConceptScore W3042765459C76155785 @default.
- W3042765459 hasConceptScore W3042765459C86803240 @default.
- W3042765459 hasConceptScore W3042765459C94915269 @default.
- W3042765459 hasConceptScore W3042765459C97541855 @default.
- W3042765459 hasOpenAccess W3042765459 @default.
- W3042765459 hasRelatedWork W2537922220 @default.
- W3042765459 hasRelatedWork W2894094671 @default.
- W3042765459 hasRelatedWork W2898290017 @default.
- W3042765459 hasRelatedWork W2908911442 @default.
- W3042765459 hasRelatedWork W2952348496 @default.
- W3042765459 hasRelatedWork W2976603810 @default.
- W3042765459 hasRelatedWork W2999490157 @default.