Matches in SemOpenAlex for { <https://semopenalex.org/work/W2807340089> ?p ?o ?g. }
- W2807340089 abstract "We introduce an approach for deep reinforcement learning (RL) that improves upon the efficiency, generalization capacity, and interpretability of conventional approaches through structured perception and relational reasoning. It uses self-attention to iteratively reason about the relations between entities in a scene and to guide a model-free policy. Our results show that in a novel navigation and planning task called Box-World, our agent finds interpretable solutions that improve upon baselines in terms of sample complexity, ability to generalize to more complex scenes than experienced during training, and overall performance. In the StarCraft II Learning Environment, our agent achieves state-of-the-art performance on six mini-games -- surpassing human grandmaster performance on four. By considering architectural inductive biases, our work opens new directions for overcoming important, but stubborn, challenges in deep RL." @default.
- W2807340089 created "2018-06-13" @default.
- W2807340089 creator A5003562101 @default.
- W2807340089 creator A5008547992 @default.
- W2807340089 creator A5031378020 @default.
- W2807340089 creator A5039679910 @default.
- W2807340089 creator A5043910056 @default.
- W2807340089 creator A5059792149 @default.
- W2807340089 creator A5061272217 @default.
- W2807340089 creator A5061984716 @default.
- W2807340089 creator A5066294254 @default.
- W2807340089 creator A5072322524 @default.
- W2807340089 creator A5074826751 @default.
- W2807340089 creator A5081026564 @default.
- W2807340089 creator A5083771180 @default.
- W2807340089 creator A5085244021 @default.
- W2807340089 creator A5088060766 @default.
- W2807340089 creator A5089917436 @default.
- W2807340089 date "2018-06-05" @default.
- W2807340089 modified "2023-10-02" @default.
- W2807340089 title "Relational Deep Reinforcement Learning." @default.
- W2807340089 cites W2519887557 @default.
- W2807340089 cites W2521274174 @default.
- W2807340089 cites W2613603362 @default.
- W2807340089 cites W2622672190 @default.
- W2807340089 cites W2624780181 @default.
- W2807340089 cites W2738675347 @default.
- W2807340089 cites W2749807327 @default.
- W2807340089 cites W2770201307 @default.
- W2807340089 cites W2775315796 @default.
- W2807340089 cites W2786036274 @default.
- W2807340089 cites W2787074649 @default.
- W2807340089 cites W2795108883 @default.
- W2807340089 cites W2797527950 @default.
- W2807340089 cites W2949267040 @default.
- W2807340089 cites W2950033033 @default.
- W2807340089 cites W2951846985 @default.
- W2807340089 cites W2952332632 @default.
- W2807340089 cites W2963403868 @default.
- W2807340089 cites W2964295739 @default.
- W2807340089 cites W2770298516 @default.
- W2807340089 hasPublicationYear "2018" @default.
- W2807340089 type Work @default.
- W2807340089 sameAs 2807340089 @default.
- W2807340089 citedByCount "82" @default.
- W2807340089 countsByYear W28073400892018 @default.
- W2807340089 countsByYear W28073400892019 @default.
- W2807340089 countsByYear W28073400892020 @default.
- W2807340089 countsByYear W28073400892021 @default.
- W2807340089 crossrefType "posted-content" @default.
- W2807340089 hasAuthorship W2807340089A5003562101 @default.
- W2807340089 hasAuthorship W2807340089A5008547992 @default.
- W2807340089 hasAuthorship W2807340089A5031378020 @default.
- W2807340089 hasAuthorship W2807340089A5039679910 @default.
- W2807340089 hasAuthorship W2807340089A5043910056 @default.
- W2807340089 hasAuthorship W2807340089A5059792149 @default.
- W2807340089 hasAuthorship W2807340089A5061272217 @default.
- W2807340089 hasAuthorship W2807340089A5061984716 @default.
- W2807340089 hasAuthorship W2807340089A5066294254 @default.
- W2807340089 hasAuthorship W2807340089A5072322524 @default.
- W2807340089 hasAuthorship W2807340089A5074826751 @default.
- W2807340089 hasAuthorship W2807340089A5081026564 @default.
- W2807340089 hasAuthorship W2807340089A5083771180 @default.
- W2807340089 hasAuthorship W2807340089A5085244021 @default.
- W2807340089 hasAuthorship W2807340089A5088060766 @default.
- W2807340089 hasAuthorship W2807340089A5089917436 @default.
- W2807340089 hasConcept C108583219 @default.
- W2807340089 hasConcept C119857082 @default.
- W2807340089 hasConcept C127413603 @default.
- W2807340089 hasConcept C134306372 @default.
- W2807340089 hasConcept C154945302 @default.
- W2807340089 hasConcept C15744967 @default.
- W2807340089 hasConcept C169760540 @default.
- W2807340089 hasConcept C177148314 @default.
- W2807340089 hasConcept C201995342 @default.
- W2807340089 hasConcept C26760741 @default.
- W2807340089 hasConcept C2780451532 @default.
- W2807340089 hasConcept C2781067378 @default.
- W2807340089 hasConcept C33923547 @default.
- W2807340089 hasConcept C41008148 @default.
- W2807340089 hasConcept C97541855 @default.
- W2807340089 hasConceptScore W2807340089C108583219 @default.
- W2807340089 hasConceptScore W2807340089C119857082 @default.
- W2807340089 hasConceptScore W2807340089C127413603 @default.
- W2807340089 hasConceptScore W2807340089C134306372 @default.
- W2807340089 hasConceptScore W2807340089C154945302 @default.
- W2807340089 hasConceptScore W2807340089C15744967 @default.
- W2807340089 hasConceptScore W2807340089C169760540 @default.
- W2807340089 hasConceptScore W2807340089C177148314 @default.
- W2807340089 hasConceptScore W2807340089C201995342 @default.
- W2807340089 hasConceptScore W2807340089C26760741 @default.
- W2807340089 hasConceptScore W2807340089C2780451532 @default.
- W2807340089 hasConceptScore W2807340089C2781067378 @default.
- W2807340089 hasConceptScore W2807340089C33923547 @default.
- W2807340089 hasConceptScore W2807340089C41008148 @default.
- W2807340089 hasConceptScore W2807340089C97541855 @default.
- W2807340089 hasLocation W28073400891 @default.
- W2807340089 hasOpenAccess W2807340089 @default.
- W2807340089 hasPrimaryLocation W28073400891 @default.
- W2807340089 hasRelatedWork W1757796397 @default.