FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2808849637> ?p ?o ?g. }

• W2808849637 abstract "Despite the remarkable success of Deep RL in learning control policies from raw pixels, the resulting models do not generalize. We demonstrate that a trained agent fails completely when facing small visual changes, and that fine-tuning---the common transfer learning paradigm---fails to adapt to these changes, to the extent that it is faster to re-train the model from scratch. We show that by separating the visual transfer task from the control policy we achieve substantially better sample efficiency and transfer behavior, allowing an agent trained on the source task to transfer well to the target tasks. The visual mapping from the target to the source domain is performed using unaligned GANs, resulting in a control policy that can be further improved using imitation learning from imperfect demonstrations. We demonstrate the approach on synthetic visual variants of the Breakout game, as well as on transfer between subsequent levels of Road Fighter, a Nintendo car-driving game. A visualization of our approach can be seen in this https URL and this https URL ." @default.
• W2808849637 created "2018-06-29" @default.
• W2808849637 creator A5028476919 @default.
• W2808849637 creator A5062045799 @default.
• W2808849637 date "2018-05-31" @default.
• W2808849637 modified "2023-09-27" @default.
• W2808849637 title "Transfer Learning for Related Reinforcement Learning Tasks via Image-to-Image Translation" @default.
• W2808849637 cites W1533861849 @default.
• W2808849637 cites W1569657508 @default.
• W2808849637 cites W1757796397 @default.
• W2808849637 cites W2099471712 @default.
• W2808849637 cites W2100495367 @default.
• W2808849637 cites W2149933564 @default.
• W2808849637 cites W2173520492 @default.
• W2808849637 cites W2186489521 @default.
• W2808849637 cites W2201581102 @default.
• W2808849637 cites W2295582178 @default.
• W2808849637 cites W2434014514 @default.
• W2808849637 cites W2552465644 @default.
• W2808849637 cites W2583761661 @default.
• W2808849637 cites W2592480533 @default.
• W2808849637 cites W2607198029 @default.
• W2808849637 cites W2618104702 @default.
• W2808849637 cites W2624780181 @default.
• W2808849637 cites W2739083961 @default.
• W2808849637 cites W2767657961 @default.
• W2808849637 cites W2770679144 @default.
• W2808849637 cites W2786036274 @default.
• W2808849637 cites W2797527950 @default.
• W2808849637 cites W2949212125 @default.
• W2808849637 cites W2951939904 @default.
• W2808849637 cites W2953326374 @default.
• W2808849637 cites W2962787969 @default.
• W2808849637 cites W2962793481 @default.
• W2808849637 cites W2962899390 @default.
• W2808849637 cites W2963363446 @default.
• W2808849637 cites W2963444790 @default.
• W2808849637 cites W2963784072 @default.
• W2808849637 cites W2964043796 @default.
• W2808849637 cites W2964062135 @default.
• W2808849637 cites W2964201809 @default.
• W2808849637 cites W2964272379 @default.
• W2808849637 cites W2426267443 @default.
• W2808849637 hasPublicationYear "2018" @default.
• W2808849637 type Work @default.
• W2808849637 sameAs 2808849637 @default.
• W2808849637 citedByCount "7" @default.
• W2808849637 countsByYear W28088496372018 @default.
• W2808849637 countsByYear W28088496372019 @default.
• W2808849637 countsByYear W28088496372020 @default.
• W2808849637 crossrefType "posted-content" @default.
• W2808849637 hasAuthorship W2808849637A5028476919 @default.
• W2808849637 hasAuthorship W2808849637A5062045799 @default.
• W2808849637 hasConcept C111919701 @default.
• W2808849637 hasConcept C119857082 @default.
• W2808849637 hasConcept C126388530 @default.
• W2808849637 hasConcept C150899416 @default.
• W2808849637 hasConcept C154945302 @default.
• W2808849637 hasConcept C15744967 @default.
• W2808849637 hasConcept C160633673 @default.
• W2808849637 hasConcept C162324750 @default.
• W2808849637 hasConcept C173608175 @default.
• W2808849637 hasConcept C187736073 @default.
• W2808849637 hasConcept C2775924081 @default.
• W2808849637 hasConcept C2776175482 @default.
• W2808849637 hasConcept C2776960227 @default.
• W2808849637 hasConcept C2780451532 @default.
• W2808849637 hasConcept C2781235140 @default.
• W2808849637 hasConcept C36464697 @default.
• W2808849637 hasConcept C41008148 @default.
• W2808849637 hasConcept C56739046 @default.
• W2808849637 hasConcept C77805123 @default.
• W2808849637 hasConcept C97541855 @default.
• W2808849637 hasConceptScore W2808849637C111919701 @default.
• W2808849637 hasConceptScore W2808849637C119857082 @default.
• W2808849637 hasConceptScore W2808849637C126388530 @default.
• W2808849637 hasConceptScore W2808849637C150899416 @default.
• W2808849637 hasConceptScore W2808849637C154945302 @default.
• W2808849637 hasConceptScore W2808849637C15744967 @default.
• W2808849637 hasConceptScore W2808849637C160633673 @default.
• W2808849637 hasConceptScore W2808849637C162324750 @default.
• W2808849637 hasConceptScore W2808849637C173608175 @default.
• W2808849637 hasConceptScore W2808849637C187736073 @default.
• W2808849637 hasConceptScore W2808849637C2775924081 @default.
• W2808849637 hasConceptScore W2808849637C2776175482 @default.
• W2808849637 hasConceptScore W2808849637C2776960227 @default.
• W2808849637 hasConceptScore W2808849637C2780451532 @default.
• W2808849637 hasConceptScore W2808849637C2781235140 @default.
• W2808849637 hasConceptScore W2808849637C36464697 @default.
• W2808849637 hasConceptScore W2808849637C41008148 @default.
• W2808849637 hasConceptScore W2808849637C56739046 @default.
• W2808849637 hasConceptScore W2808849637C77805123 @default.
• W2808849637 hasConceptScore W2808849637C97541855 @default.
• W2808849637 hasLocation W28088496371 @default.
• W2808849637 hasOpenAccess W2808849637 @default.
• W2808849637 hasPrimaryLocation W28088496371 @default.
• W2808849637 hasRelatedWork W2604299654 @default.
• W2808849637 hasRelatedWork W2741926431 @default.
• W2808849637 hasRelatedWork W2914688076 @default.
• W2808849637 hasRelatedWork W2949121148 @default.

• next