Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288094104> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4288094104 abstract "Humans are masters at quickly learning many complex tasks, relying on an approximate understanding of the dynamics of their environments. In much the same way, we would like our learning agents to quickly adapt to new tasks. In this paper, we explore how model-based Reinforcement Learning (RL) can facilitate transfer to new tasks. We develop an algorithm that learns an action-conditional, predictive model of expected future observations, rewards and values from which a policy can be derived by following the gradient of the estimated value along imagined trajectories. We show how robust policy optimization can be achieved in robot manipulation tasks even with approximate models that are learned directly from vision and proprioception. We evaluate the efficacy of our approach in a transfer learning scenario, re-using previously learned models on tasks with different reward structures and visual distractors, and show a significant improvement in learning speed compared to strong off-policy baselines. Videos with results can be found at https://sites.google.com/view/ivg-corl19" @default.
- W4288094104 created "2022-07-28" @default.
- W4288094104 creator A5007133617 @default.
- W4288094104 creator A5017985443 @default.
- W4288094104 creator A5018196238 @default.
- W4288094104 creator A5037305533 @default.
- W4288094104 creator A5041323275 @default.
- W4288094104 creator A5053312475 @default.
- W4288094104 creator A5054636066 @default.
- W4288094104 creator A5062951341 @default.
- W4288094104 creator A5065489996 @default.
- W4288094104 date "2019-10-09" @default.
- W4288094104 modified "2023-09-25" @default.
- W4288094104 title "Imagined Value Gradients: Model-Based Policy Optimization with Transferable Latent Dynamics Models" @default.
- W4288094104 doi "https://doi.org/10.48550/arxiv.1910.04142" @default.
- W4288094104 hasPublicationYear "2019" @default.
- W4288094104 type Work @default.
- W4288094104 citedByCount "0" @default.
- W4288094104 crossrefType "posted-content" @default.
- W4288094104 hasAuthorship W4288094104A5007133617 @default.
- W4288094104 hasAuthorship W4288094104A5017985443 @default.
- W4288094104 hasAuthorship W4288094104A5018196238 @default.
- W4288094104 hasAuthorship W4288094104A5037305533 @default.
- W4288094104 hasAuthorship W4288094104A5041323275 @default.
- W4288094104 hasAuthorship W4288094104A5053312475 @default.
- W4288094104 hasAuthorship W4288094104A5054636066 @default.
- W4288094104 hasAuthorship W4288094104A5062951341 @default.
- W4288094104 hasAuthorship W4288094104A5065489996 @default.
- W4288094104 hasBestOaLocation W42880941041 @default.
- W4288094104 hasConcept C119857082 @default.
- W4288094104 hasConcept C121332964 @default.
- W4288094104 hasConcept C145912823 @default.
- W4288094104 hasConcept C150899416 @default.
- W4288094104 hasConcept C154945302 @default.
- W4288094104 hasConcept C15744967 @default.
- W4288094104 hasConcept C19417346 @default.
- W4288094104 hasConcept C2776291640 @default.
- W4288094104 hasConcept C2779436431 @default.
- W4288094104 hasConcept C2780791683 @default.
- W4288094104 hasConcept C41008148 @default.
- W4288094104 hasConcept C62520636 @default.
- W4288094104 hasConcept C97541855 @default.
- W4288094104 hasConceptScore W4288094104C119857082 @default.
- W4288094104 hasConceptScore W4288094104C121332964 @default.
- W4288094104 hasConceptScore W4288094104C145912823 @default.
- W4288094104 hasConceptScore W4288094104C150899416 @default.
- W4288094104 hasConceptScore W4288094104C154945302 @default.
- W4288094104 hasConceptScore W4288094104C15744967 @default.
- W4288094104 hasConceptScore W4288094104C19417346 @default.
- W4288094104 hasConceptScore W4288094104C2776291640 @default.
- W4288094104 hasConceptScore W4288094104C2779436431 @default.
- W4288094104 hasConceptScore W4288094104C2780791683 @default.
- W4288094104 hasConceptScore W4288094104C41008148 @default.
- W4288094104 hasConceptScore W4288094104C62520636 @default.
- W4288094104 hasConceptScore W4288094104C97541855 @default.
- W4288094104 hasLocation W42880941041 @default.
- W4288094104 hasOpenAccess W4288094104 @default.
- W4288094104 hasPrimaryLocation W42880941041 @default.
- W4288094104 hasRelatedWork W1997664188 @default.
- W4288094104 hasRelatedWork W2960456850 @default.
- W4288094104 hasRelatedWork W3022038857 @default.
- W4288094104 hasRelatedWork W4281382123 @default.
- W4288094104 hasRelatedWork W4281645081 @default.
- W4288094104 hasRelatedWork W4294306704 @default.
- W4288094104 hasRelatedWork W4308262314 @default.
- W4288094104 hasRelatedWork W4318834068 @default.
- W4288094104 hasRelatedWork W4318957922 @default.
- W4288094104 hasRelatedWork W4319083788 @default.
- W4288094104 isParatext "false" @default.
- W4288094104 isRetracted "false" @default.
- W4288094104 workType "article" @default.