Matches in SemOpenAlex for { <https://semopenalex.org/work/W2753847160> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W2753847160 endingPage "3338" @default.
- W2753847160 startingPage "3329" @default.
- W2753847160 abstract "Reinforcement learning (RL) in partially observable settings is challenging because the agent’s immediate observations are not Markov. Recently proposed methods can learn variable-order Markov models of the underlying process but have steep memory requirements and are sensitive to aliasing between observation histories due to sensor noise. This paper proposes utile context tree weighting (UCTW), a model-learning method that addresses these limitations. UCTW dynamically expands a suffix tree while ensuring that the total size of the model, but not its depth, remains bounded. We show that UCTW approximately matches the performance of state-of-the-art alternatives at stochastic time-series prediction while using at least an order of magnitude less memory. We also apply UCTW to model-based RL, showing that, on tasks that require memory of past observations, UCTW can learn without prior knowledge of a good state representation, or even the length of history upon which such a representation should depend." @default.
- W2753847160 created "2017-09-15" @default.
- W2753847160 creator A5056879203 @default.
- W2753847160 creator A5076409021 @default.
- W2753847160 date "2017-01-01" @default.
- W2753847160 modified "2023-10-18" @default.
- W2753847160 title "Utile Context Tree Weighting" @default.
- W2753847160 hasPublicationYear "2017" @default.
- W2753847160 type Work @default.
- W2753847160 sameAs 2753847160 @default.
- W2753847160 citedByCount "0" @default.
- W2753847160 crossrefType "proceedings-article" @default.
- W2753847160 hasAuthorship W2753847160A5056879203 @default.
- W2753847160 hasAuthorship W2753847160A5076409021 @default.
- W2753847160 hasConcept C105795698 @default.
- W2753847160 hasConcept C113174947 @default.
- W2753847160 hasConcept C11413529 @default.
- W2753847160 hasConcept C119857082 @default.
- W2753847160 hasConcept C126838900 @default.
- W2753847160 hasConcept C134306372 @default.
- W2753847160 hasConcept C136536468 @default.
- W2753847160 hasConcept C151730666 @default.
- W2753847160 hasConcept C154945302 @default.
- W2753847160 hasConcept C159886148 @default.
- W2753847160 hasConcept C17744445 @default.
- W2753847160 hasConcept C183115368 @default.
- W2753847160 hasConcept C199539241 @default.
- W2753847160 hasConcept C23224414 @default.
- W2753847160 hasConcept C2776359362 @default.
- W2753847160 hasConcept C2779343474 @default.
- W2753847160 hasConcept C33923547 @default.
- W2753847160 hasConcept C4069607 @default.
- W2753847160 hasConcept C41008148 @default.
- W2753847160 hasConcept C71924100 @default.
- W2753847160 hasConcept C86803240 @default.
- W2753847160 hasConcept C94625758 @default.
- W2753847160 hasConcept C97541855 @default.
- W2753847160 hasConceptScore W2753847160C105795698 @default.
- W2753847160 hasConceptScore W2753847160C113174947 @default.
- W2753847160 hasConceptScore W2753847160C11413529 @default.
- W2753847160 hasConceptScore W2753847160C119857082 @default.
- W2753847160 hasConceptScore W2753847160C126838900 @default.
- W2753847160 hasConceptScore W2753847160C134306372 @default.
- W2753847160 hasConceptScore W2753847160C136536468 @default.
- W2753847160 hasConceptScore W2753847160C151730666 @default.
- W2753847160 hasConceptScore W2753847160C154945302 @default.
- W2753847160 hasConceptScore W2753847160C159886148 @default.
- W2753847160 hasConceptScore W2753847160C17744445 @default.
- W2753847160 hasConceptScore W2753847160C183115368 @default.
- W2753847160 hasConceptScore W2753847160C199539241 @default.
- W2753847160 hasConceptScore W2753847160C23224414 @default.
- W2753847160 hasConceptScore W2753847160C2776359362 @default.
- W2753847160 hasConceptScore W2753847160C2779343474 @default.
- W2753847160 hasConceptScore W2753847160C33923547 @default.
- W2753847160 hasConceptScore W2753847160C4069607 @default.
- W2753847160 hasConceptScore W2753847160C41008148 @default.
- W2753847160 hasConceptScore W2753847160C71924100 @default.
- W2753847160 hasConceptScore W2753847160C86803240 @default.
- W2753847160 hasConceptScore W2753847160C94625758 @default.
- W2753847160 hasConceptScore W2753847160C97541855 @default.
- W2753847160 hasLocation W27538471601 @default.
- W2753847160 hasOpenAccess W2753847160 @default.
- W2753847160 hasPrimaryLocation W27538471601 @default.
- W2753847160 hasRelatedWork W1554774506 @default.
- W2753847160 hasRelatedWork W175242119 @default.
- W2753847160 hasRelatedWork W2031063962 @default.
- W2753847160 hasRelatedWork W2157457978 @default.
- W2753847160 hasRelatedWork W2289584534 @default.
- W2753847160 hasRelatedWork W2616944917 @default.
- W2753847160 hasRelatedWork W2767624189 @default.
- W2753847160 hasRelatedWork W2787477569 @default.
- W2753847160 hasRelatedWork W2796054047 @default.
- W2753847160 hasRelatedWork W2948204052 @default.
- W2753847160 hasRelatedWork W2949238434 @default.
- W2753847160 hasRelatedWork W2951637760 @default.
- W2753847160 hasRelatedWork W3007247345 @default.
- W2753847160 hasRelatedWork W3097059970 @default.
- W2753847160 hasRelatedWork W3106325613 @default.
- W2753847160 hasRelatedWork W3208550664 @default.
- W2753847160 hasRelatedWork W3214059887 @default.
- W2753847160 hasRelatedWork W50830905 @default.
- W2753847160 hasRelatedWork W605299978 @default.
- W2753847160 hasRelatedWork W3124153248 @default.
- W2753847160 isParatext "false" @default.
- W2753847160 isRetracted "false" @default.
- W2753847160 magId "2753847160" @default.
- W2753847160 workType "article" @default.