FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3090370023> ?p ?o ?g. }

• W3090370023 abstract "Exploration and credit assignment under sparse rewards are still challenging problems. We argue that these challenges arise in part due to the intrinsic rigidity of operating at the level of actions. Actions can precisely define how to perform an activity but are ill-suited to describe what activity to perform. Instead, causal effects are inherently composable and temporally abstract, making them ideal for descriptive tasks. By leveraging a hierarchy of causal effects, this study aims to expedite the learning of task-specific behavior and aid exploration. Borrowing counterfactual and normality measures from causal literature, we disentangle controllable effects from effects caused by other dynamics of the environment. We propose CEHRL, a hierarchical method that models the distribution of controllable effects using a Variational Autoencoder. This distribution is used by a high-level policy to 1) explore the environment via random effect exploration so that novel effects are continuously discovered and learned, and to 2) learn task-specific behavior by prioritizing the effects that maximize a given reward function. In comparison to exploring with random actions, experimental results show that random effect exploration is a more efficient mechanism and that by assigning credit to few effects rather than many actions, CEHRL learns tasks more rapidly." @default.
• W3090370023 created "2020-10-08" @default.
• W3090370023 creator A5038506987 @default.
• W3090370023 creator A5056574222 @default.
• W3090370023 date "2020-10-03" @default.
• W3090370023 modified "2023-10-18" @default.
• W3090370023 title "Disentangling causal effects for hierarchical reinforcement learning." @default.
• W3090370023 cites W1515851193 @default.
• W3090370023 cites W1757796397 @default.
• W3090370023 cites W1983497409 @default.
• W3090370023 cites W2011861615 @default.
• W3090370023 cites W2109910161 @default.
• W3090370023 cites W2139590055 @default.
• W3090370023 cites W2139612737 @default.
• W3090370023 cites W2143891888 @default.
• W3090370023 cites W2335959470 @default.
• W3090370023 cites W2523728418 @default.
• W3090370023 cites W2585464547 @default.
• W3090370023 cites W2606433045 @default.
• W3090370023 cites W2614839826 @default.
• W3090370023 cites W2733961795 @default.
• W3090370023 cites W2786928559 @default.
• W3090370023 cites W2788741142 @default.
• W3090370023 cites W2823112946 @default.
• W3090370023 cites W2899205164 @default.
• W3090370023 cites W2900677074 @default.
• W3090370023 cites W2909534157 @default.
• W3090370023 cites W2911448865 @default.
• W3090370023 cites W2912031212 @default.
• W3090370023 cites W2912692476 @default.
• W3090370023 cites W2914351253 @default.
• W3090370023 cites W2914607694 @default.
• W3090370023 cites W2924740141 @default.
• W3090370023 cites W2943863006 @default.
• W3090370023 cites W2949267040 @default.
• W3090370023 cites W2951004968 @default.
• W3090370023 cites W2951799221 @default.
• W3090370023 cites W2954142106 @default.
• W3090370023 cites W2954360742 @default.
• W3090370023 cites W2963126744 @default.
• W3090370023 cites W2963293881 @default.
• W3090370023 cites W2963477884 @default.
• W3090370023 cites W2971040589 @default.
• W3090370023 cites W2973525135 @default.
• W3090370023 cites W2979174981 @default.
• W3090370023 cites W2979225831 @default.
• W3090370023 cites W2996695841 @default.
• W3090370023 cites W2973223114 @default.
• W3090370023 hasPublicationYear "2020" @default.
• W3090370023 type Work @default.
• W3090370023 sameAs 3090370023 @default.
• W3090370023 citedByCount "2" @default.
• W3090370023 countsByYear W30903700232021 @default.
• W3090370023 crossrefType "posted-content" @default.
• W3090370023 hasAuthorship W3090370023A5038506987 @default.
• W3090370023 hasAuthorship W3090370023A5056574222 @default.
• W3090370023 hasConcept C101738243 @default.
• W3090370023 hasConcept C108650721 @default.
• W3090370023 hasConcept C119857082 @default.
• W3090370023 hasConcept C154945302 @default.
• W3090370023 hasConcept C15744967 @default.
• W3090370023 hasConcept C162324750 @default.
• W3090370023 hasConcept C187736073 @default.
• W3090370023 hasConcept C2780451532 @default.
• W3090370023 hasConcept C31170391 @default.
• W3090370023 hasConcept C34447519 @default.
• W3090370023 hasConcept C41008148 @default.
• W3090370023 hasConcept C50644808 @default.
• W3090370023 hasConcept C77805123 @default.
• W3090370023 hasConcept C97541855 @default.
• W3090370023 hasConceptScore W3090370023C101738243 @default.
• W3090370023 hasConceptScore W3090370023C108650721 @default.
• W3090370023 hasConceptScore W3090370023C119857082 @default.
• W3090370023 hasConceptScore W3090370023C154945302 @default.
• W3090370023 hasConceptScore W3090370023C15744967 @default.
• W3090370023 hasConceptScore W3090370023C162324750 @default.
• W3090370023 hasConceptScore W3090370023C187736073 @default.
• W3090370023 hasConceptScore W3090370023C2780451532 @default.
• W3090370023 hasConceptScore W3090370023C31170391 @default.
• W3090370023 hasConceptScore W3090370023C34447519 @default.
• W3090370023 hasConceptScore W3090370023C41008148 @default.
• W3090370023 hasConceptScore W3090370023C50644808 @default.
• W3090370023 hasConceptScore W3090370023C77805123 @default.
• W3090370023 hasConceptScore W3090370023C97541855 @default.
• W3090370023 hasLocation W30903700231 @default.
• W3090370023 hasOpenAccess W3090370023 @default.
• W3090370023 hasPrimaryLocation W30903700231 @default.
• W3090370023 hasRelatedWork W1519539863 @default.
• W3090370023 hasRelatedWork W1607218107 @default.
• W3090370023 hasRelatedWork W2028496760 @default.
• W3090370023 hasRelatedWork W2078150668 @default.
• W3090370023 hasRelatedWork W2604626881 @default.
• W3090370023 hasRelatedWork W2619583400 @default.
• W3090370023 hasRelatedWork W2883343497 @default.
• W3090370023 hasRelatedWork W2892266804 @default.
• W3090370023 hasRelatedWork W2949969799 @default.
• W3090370023 hasRelatedWork W2972522277 @default.
• W3090370023 hasRelatedWork W2972871590 @default.
• W3090370023 hasRelatedWork W2995104119 @default.
• W3090370023 hasRelatedWork W2998135952 @default.

• next