Matches in SemOpenAlex for { <https://semopenalex.org/work/W2804589917> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W2804589917 abstract "We devise a distributional variant of gradient temporal-difference (TD) learning. Distributional reinforcement learning has been demonstrated to outperform the regular one in the recent study citep{bellemare2017distributional}. In the policy evaluation setting, we design two new algorithms called distributional GTD2 and distributional TDC using the Cram{'e}r distance on the distributional version of the Bellman error objective function, which inherits advantages of both the nonlinear gradient TD algorithms and the distributional RL approach. In the control setting, we propose the distributional Greedy-GQ using the similar derivation. We prove the asymptotic almost-sure convergence of distributional GTD2 and TDC to a local optimal solution for general smooth function approximators, which includes neural networks that have been widely used in recent study to solve the real-life RL problems. In each step, the computational complexities of above three algorithms are linear w.r.t. the number of the parameters of the function approximator, thus can be implemented efficiently for neural networks." @default.
- W2804589917 created "2018-06-01" @default.
- W2804589917 creator A5036260775 @default.
- W2804589917 creator A5061203282 @default.
- W2804589917 creator A5083934553 @default.
- W2804589917 date "2018-05-20" @default.
- W2804589917 modified "2023-10-18" @default.
- W2804589917 title "Nonlinear Distributional Gradient Temporal-Difference Learning" @default.
- W2804589917 cites W13294968 @default.
- W2804589917 cites W134786152 @default.
- W2804589917 cites W1564755532 @default.
- W2804589917 cites W1646707810 @default.
- W2804589917 cites W1663973292 @default.
- W2804589917 cites W2006903949 @default.
- W2804589917 cites W2071983464 @default.
- W2804589917 cites W2075268401 @default.
- W2804589917 cites W2112663282 @default.
- W2804589917 cites W2121703796 @default.
- W2804589917 cites W2134042548 @default.
- W2804589917 cites W2139418546 @default.
- W2804589917 cites W2145339207 @default.
- W2804589917 cites W2153678894 @default.
- W2804589917 cites W2155968351 @default.
- W2804589917 cites W2156974606 @default.
- W2804589917 cites W2165905123 @default.
- W2804589917 cites W2279759792 @default.
- W2804589917 cites W2489939061 @default.
- W2804589917 cites W2619903301 @default.
- W2804589917 cites W2755611049 @default.
- W2804589917 cites W2765302304 @default.
- W2804589917 cites W2963423916 @default.
- W2804589917 cites W2963610502 @default.
- W2804589917 cites W2963871073 @default.
- W2804589917 doi "https://doi.org/10.48550/arxiv.1805.07732" @default.
- W2804589917 hasPublicationYear "2018" @default.
- W2804589917 type Work @default.
- W2804589917 sameAs 2804589917 @default.
- W2804589917 citedByCount "4" @default.
- W2804589917 countsByYear W28045899172019 @default.
- W2804589917 countsByYear W28045899172020 @default.
- W2804589917 crossrefType "posted-content" @default.
- W2804589917 hasAuthorship W2804589917A5036260775 @default.
- W2804589917 hasAuthorship W2804589917A5061203282 @default.
- W2804589917 hasAuthorship W2804589917A5083934553 @default.
- W2804589917 hasBestOaLocation W28045899171 @default.
- W2804589917 hasConcept C11413529 @default.
- W2804589917 hasConcept C121332964 @default.
- W2804589917 hasConcept C126255220 @default.
- W2804589917 hasConcept C14036430 @default.
- W2804589917 hasConcept C14646407 @default.
- W2804589917 hasConcept C154945302 @default.
- W2804589917 hasConcept C158622935 @default.
- W2804589917 hasConcept C162324750 @default.
- W2804589917 hasConcept C196340769 @default.
- W2804589917 hasConcept C2777303404 @default.
- W2804589917 hasConcept C28826006 @default.
- W2804589917 hasConcept C33923547 @default.
- W2804589917 hasConcept C41008148 @default.
- W2804589917 hasConcept C50522688 @default.
- W2804589917 hasConcept C50644808 @default.
- W2804589917 hasConcept C62520636 @default.
- W2804589917 hasConcept C78458016 @default.
- W2804589917 hasConcept C86803240 @default.
- W2804589917 hasConcept C97541855 @default.
- W2804589917 hasConceptScore W2804589917C11413529 @default.
- W2804589917 hasConceptScore W2804589917C121332964 @default.
- W2804589917 hasConceptScore W2804589917C126255220 @default.
- W2804589917 hasConceptScore W2804589917C14036430 @default.
- W2804589917 hasConceptScore W2804589917C14646407 @default.
- W2804589917 hasConceptScore W2804589917C154945302 @default.
- W2804589917 hasConceptScore W2804589917C158622935 @default.
- W2804589917 hasConceptScore W2804589917C162324750 @default.
- W2804589917 hasConceptScore W2804589917C196340769 @default.
- W2804589917 hasConceptScore W2804589917C2777303404 @default.
- W2804589917 hasConceptScore W2804589917C28826006 @default.
- W2804589917 hasConceptScore W2804589917C33923547 @default.
- W2804589917 hasConceptScore W2804589917C41008148 @default.
- W2804589917 hasConceptScore W2804589917C50522688 @default.
- W2804589917 hasConceptScore W2804589917C50644808 @default.
- W2804589917 hasConceptScore W2804589917C62520636 @default.
- W2804589917 hasConceptScore W2804589917C78458016 @default.
- W2804589917 hasConceptScore W2804589917C86803240 @default.
- W2804589917 hasConceptScore W2804589917C97541855 @default.
- W2804589917 hasLocation W28045899171 @default.
- W2804589917 hasOpenAccess W2804589917 @default.
- W2804589917 hasPrimaryLocation W28045899171 @default.
- W2804589917 hasRelatedWork W2129532790 @default.
- W2804589917 hasRelatedWork W2145363145 @default.
- W2804589917 hasRelatedWork W2149418961 @default.
- W2804589917 hasRelatedWork W2151568021 @default.
- W2804589917 hasRelatedWork W2156232164 @default.
- W2804589917 hasRelatedWork W2341346307 @default.
- W2804589917 hasRelatedWork W2808418668 @default.
- W2804589917 hasRelatedWork W3023701936 @default.
- W2804589917 hasRelatedWork W4289355352 @default.
- W2804589917 hasRelatedWork W4308702637 @default.
- W2804589917 isParatext "false" @default.
- W2804589917 isRetracted "false" @default.
- W2804589917 magId "2804589917" @default.
- W2804589917 workType "article" @default.