Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387171566> ?p ?o ?g. }
Showing items 1 to 51 of
51
with 100 items per page.
- W4387171566 abstract "Learning fair policies in reinforcement learning (RL) is important when the RL agent may impact many users. We investigate a variant of this problem where equity is still desired, but some users may be entitled to preferential treatment. In this paper, we formalize this more sophisticated fair optimization problem in deep RL using generalized fair social welfare functions (SWF), provide a theoretical discussion to justify our approach, explain how deep RL algorithms can be adapted to tackle it, and empirically validate our propositions on several domains. Our contributions are both theoretical and algorithmic, notably: (1) We obtain a general bound on the suboptimality gap in terms of SWF-optimality using average reward of a policy SWF-optimal for the discounted reward, which notably justifies using standard deep RL algorithms, even for the average reward; (2) Our algorithmic innovations include a state-augmented DQN-based method for learning either deterministic or stochastic policies, which also applies to the usual fair optimization setting without any preferential treatment." @default.
- W4387171566 created "2023-09-30" @default.
- W4387171566 creator A5029379398 @default.
- W4387171566 creator A5073106112 @default.
- W4387171566 creator A5087259735 @default.
- W4387171566 date "2023-09-28" @default.
- W4387171566 modified "2023-09-30" @default.
- W4387171566 title "Fair Deep Reinforcement Learning with Preferential Treatment" @default.
- W4387171566 doi "https://doi.org/10.3233/faia230606" @default.
- W4387171566 hasPublicationYear "2023" @default.
- W4387171566 type Work @default.
- W4387171566 citedByCount "0" @default.
- W4387171566 crossrefType "book-chapter" @default.
- W4387171566 hasAuthorship W4387171566A5029379398 @default.
- W4387171566 hasAuthorship W4387171566A5073106112 @default.
- W4387171566 hasAuthorship W4387171566A5087259735 @default.
- W4387171566 hasBestOaLocation W43871715661 @default.
- W4387171566 hasConcept C108583219 @default.
- W4387171566 hasConcept C126255220 @default.
- W4387171566 hasConcept C154945302 @default.
- W4387171566 hasConcept C17744445 @default.
- W4387171566 hasConcept C199539241 @default.
- W4387171566 hasConcept C199728807 @default.
- W4387171566 hasConcept C33923547 @default.
- W4387171566 hasConcept C41008148 @default.
- W4387171566 hasConcept C97541855 @default.
- W4387171566 hasConceptScore W4387171566C108583219 @default.
- W4387171566 hasConceptScore W4387171566C126255220 @default.
- W4387171566 hasConceptScore W4387171566C154945302 @default.
- W4387171566 hasConceptScore W4387171566C17744445 @default.
- W4387171566 hasConceptScore W4387171566C199539241 @default.
- W4387171566 hasConceptScore W4387171566C199728807 @default.
- W4387171566 hasConceptScore W4387171566C33923547 @default.
- W4387171566 hasConceptScore W4387171566C41008148 @default.
- W4387171566 hasConceptScore W4387171566C97541855 @default.
- W4387171566 hasLocation W43871715661 @default.
- W4387171566 hasOpenAccess W4387171566 @default.
- W4387171566 hasPrimaryLocation W43871715661 @default.
- W4387171566 hasRelatedWork W260766989 @default.
- W4387171566 hasRelatedWork W2731899572 @default.
- W4387171566 hasRelatedWork W2939353110 @default.
- W4387171566 hasRelatedWork W2959276766 @default.
- W4387171566 hasRelatedWork W3009238340 @default.
- W4387171566 hasRelatedWork W3074294383 @default.
- W4387171566 hasRelatedWork W3139193008 @default.
- W4387171566 hasRelatedWork W3215138031 @default.
- W4387171566 hasRelatedWork W4206669594 @default.
- W4387171566 hasRelatedWork W4295941380 @default.
- W4387171566 isParatext "false" @default.
- W4387171566 isRetracted "false" @default.
- W4387171566 workType "book-chapter" @default.