Matches in SemOpenAlex for { <https://semopenalex.org/work/W3173921802> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W3173921802 endingPage "347" @default.
- W3173921802 startingPage "336" @default.
- W3173921802 abstract "This paper focuses on finding reinforcement learning policies for control systems with hard state and action constraints. Despite its success in many domains, reinforcement learning is challenging to apply to problems with hard constraints, especially if both the state variables and actions are constrained. Previous works seeking to ensure constraint satisfaction, or safety, have focused on adding a projection step to a learned policy. Yet, this approach requires solving an optimization problem at every policy execution step, which can lead to significant computational costs. To tackle this problem, this paper proposes a new approach, termed Vertex Networks (VNs), with guarantees on safety during exploration and on learned control policies by incorporating the safety constraints into the policy network architecture. Leveraging the geometric property that all points within a convex set can be represented as the convex combination of its vertices, the proposed algorithm first learns the convex combination weights and then uses these weights along with the pre-calculated vertices to output an action. The output action is guaranteed to be safe by construction. Numerical examples illustrate that the proposed VN algorithm outperforms vanilla reinforcement learning in a variety of benchmark control tasks." @default.
- W3173921802 created "2021-07-05" @default.
- W3173921802 creator A5008161296 @default.
- W3173921802 creator A5011436134 @default.
- W3173921802 creator A5013901541 @default.
- W3173921802 creator A5055691206 @default.
- W3173921802 date "2021-05-29" @default.
- W3173921802 modified "2023-09-28" @default.
- W3173921802 title "Safe Reinforcement Learning of Control-Affine Systems with Vertex Networks." @default.
- W3173921802 hasPublicationYear "2021" @default.
- W3173921802 type Work @default.
- W3173921802 sameAs 3173921802 @default.
- W3173921802 citedByCount "0" @default.
- W3173921802 crossrefType "journal-article" @default.
- W3173921802 hasAuthorship W3173921802A5008161296 @default.
- W3173921802 hasAuthorship W3173921802A5011436134 @default.
- W3173921802 hasAuthorship W3173921802A5013901541 @default.
- W3173921802 hasAuthorship W3173921802A5055691206 @default.
- W3173921802 hasConcept C126255220 @default.
- W3173921802 hasConcept C13280743 @default.
- W3173921802 hasConcept C154945302 @default.
- W3173921802 hasConcept C177264268 @default.
- W3173921802 hasConcept C185798385 @default.
- W3173921802 hasConcept C199360897 @default.
- W3173921802 hasConcept C205649164 @default.
- W3173921802 hasConcept C33923547 @default.
- W3173921802 hasConcept C41008148 @default.
- W3173921802 hasConcept C97541855 @default.
- W3173921802 hasConceptScore W3173921802C126255220 @default.
- W3173921802 hasConceptScore W3173921802C13280743 @default.
- W3173921802 hasConceptScore W3173921802C154945302 @default.
- W3173921802 hasConceptScore W3173921802C177264268 @default.
- W3173921802 hasConceptScore W3173921802C185798385 @default.
- W3173921802 hasConceptScore W3173921802C199360897 @default.
- W3173921802 hasConceptScore W3173921802C205649164 @default.
- W3173921802 hasConceptScore W3173921802C33923547 @default.
- W3173921802 hasConceptScore W3173921802C41008148 @default.
- W3173921802 hasConceptScore W3173921802C97541855 @default.
- W3173921802 hasLocation W31739218021 @default.
- W3173921802 hasOpenAccess W3173921802 @default.
- W3173921802 hasPrimaryLocation W31739218021 @default.
- W3173921802 hasRelatedWork W1577409703 @default.
- W3173921802 hasRelatedWork W1598210495 @default.
- W3173921802 hasRelatedWork W2042241295 @default.
- W3173921802 hasRelatedWork W2075632665 @default.
- W3173921802 hasRelatedWork W2107166738 @default.
- W3173921802 hasRelatedWork W2558997757 @default.
- W3173921802 hasRelatedWork W2576280707 @default.
- W3173921802 hasRelatedWork W2585185247 @default.
- W3173921802 hasRelatedWork W2962734844 @default.
- W3173921802 hasRelatedWork W2980693811 @default.
- W3173921802 hasRelatedWork W3001024844 @default.
- W3173921802 hasRelatedWork W3005774977 @default.
- W3173921802 hasRelatedWork W3012839911 @default.
- W3173921802 hasRelatedWork W3028311185 @default.
- W3173921802 hasRelatedWork W3091364228 @default.
- W3173921802 hasRelatedWork W3124254895 @default.
- W3173921802 hasRelatedWork W3126230687 @default.
- W3173921802 hasRelatedWork W3155204957 @default.
- W3173921802 hasRelatedWork W3172490820 @default.
- W3173921802 hasRelatedWork W3188777742 @default.
- W3173921802 isParatext "false" @default.
- W3173921802 isRetracted "false" @default.
- W3173921802 magId "3173921802" @default.
- W3173921802 workType "article" @default.