Matches in SemOpenAlex for { <https://semopenalex.org/work/W2145756561> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2145756561 endingPage "1011" @default.
- W2145756561 startingPage "1005" @default.
- W2145756561 abstract "Dynamic Programming, Q-learning and other discrete Markov Decision Process solvers can be applied to continuous d-dimensional state-spaces by quantizing the state space into an array of boxes. This is often problematic above two dimensions: a coarse quantization can lead to poor policies, and fine quantization is too expensive. Possible solutions are variable-resolution discretization, or function approximation by neural nets. A third option, which has been little studied in the reinforcement learning literature, is interpolation on a coarse grid. In this paper we study interpolation techniques that can result in vast improvements in the online behavior of the resulting control systems: multilinear interpolation, and an interpolation algorithm based on an interesting regular triangulation of d-dimensional space. We adapt these interpolators under three reinforcement learning paradigms: (i) offline value iteration with a known model, (ii) Q-learning, and (iii) online value iteration with a previously unknown model learned from data. We describe empirical results, and the resulting implications for practical learning of continuous non-linear dynamic control." @default.
- W2145756561 created "2016-06-24" @default.
- W2145756561 creator A5047831357 @default.
- W2145756561 date "1996-12-03" @default.
- W2145756561 modified "2023-09-24" @default.
- W2145756561 title "Multidimensional Triangulation and Interpolation for Reinforcement Learning" @default.
- W2145756561 cites W1547105496 @default.
- W2145756561 cites W1776040741 @default.
- W2145756561 cites W1966195676 @default.
- W2145756561 cites W2035446426 @default.
- W2145756561 cites W2048226872 @default.
- W2145756561 cites W2100677568 @default.
- W2145756561 cites W2112663282 @default.
- W2145756561 cites W2117341272 @default.
- W2145756561 cites W2124175081 @default.
- W2145756561 cites W2125074935 @default.
- W2145756561 cites W3011120880 @default.
- W2145756561 hasPublicationYear "1996" @default.
- W2145756561 type Work @default.
- W2145756561 sameAs 2145756561 @default.
- W2145756561 citedByCount "29" @default.
- W2145756561 countsByYear W21457565612012 @default.
- W2145756561 countsByYear W21457565612013 @default.
- W2145756561 countsByYear W21457565612014 @default.
- W2145756561 countsByYear W21457565612015 @default.
- W2145756561 countsByYear W21457565612016 @default.
- W2145756561 countsByYear W21457565612017 @default.
- W2145756561 countsByYear W21457565612019 @default.
- W2145756561 countsByYear W21457565612020 @default.
- W2145756561 countsByYear W21457565612021 @default.
- W2145756561 crossrefType "proceedings-article" @default.
- W2145756561 hasAuthorship W2145756561A5047831357 @default.
- W2145756561 hasConcept C105795698 @default.
- W2145756561 hasConcept C106189395 @default.
- W2145756561 hasConcept C11413529 @default.
- W2145756561 hasConcept C115961682 @default.
- W2145756561 hasConcept C126255220 @default.
- W2145756561 hasConcept C134306372 @default.
- W2145756561 hasConcept C137800194 @default.
- W2145756561 hasConcept C14646407 @default.
- W2145756561 hasConcept C153180895 @default.
- W2145756561 hasConcept C154945302 @default.
- W2145756561 hasConcept C159886148 @default.
- W2145756561 hasConcept C171836373 @default.
- W2145756561 hasConcept C28855332 @default.
- W2145756561 hasConcept C33923547 @default.
- W2145756561 hasConcept C41008148 @default.
- W2145756561 hasConcept C73000952 @default.
- W2145756561 hasConcept C97541855 @default.
- W2145756561 hasConceptScore W2145756561C105795698 @default.
- W2145756561 hasConceptScore W2145756561C106189395 @default.
- W2145756561 hasConceptScore W2145756561C11413529 @default.
- W2145756561 hasConceptScore W2145756561C115961682 @default.
- W2145756561 hasConceptScore W2145756561C126255220 @default.
- W2145756561 hasConceptScore W2145756561C134306372 @default.
- W2145756561 hasConceptScore W2145756561C137800194 @default.
- W2145756561 hasConceptScore W2145756561C14646407 @default.
- W2145756561 hasConceptScore W2145756561C153180895 @default.
- W2145756561 hasConceptScore W2145756561C154945302 @default.
- W2145756561 hasConceptScore W2145756561C159886148 @default.
- W2145756561 hasConceptScore W2145756561C171836373 @default.
- W2145756561 hasConceptScore W2145756561C28855332 @default.
- W2145756561 hasConceptScore W2145756561C33923547 @default.
- W2145756561 hasConceptScore W2145756561C41008148 @default.
- W2145756561 hasConceptScore W2145756561C73000952 @default.
- W2145756561 hasConceptScore W2145756561C97541855 @default.
- W2145756561 hasLocation W21457565611 @default.
- W2145756561 hasOpenAccess W2145756561 @default.
- W2145756561 hasPrimaryLocation W21457565611 @default.
- W2145756561 hasRelatedWork W1547105496 @default.
- W2145756561 hasRelatedWork W1552830313 @default.
- W2145756561 hasRelatedWork W1576452626 @default.
- W2145756561 hasRelatedWork W1601081659 @default.
- W2145756561 hasRelatedWork W1759147280 @default.
- W2145756561 hasRelatedWork W1776040741 @default.
- W2145756561 hasRelatedWork W1966195676 @default.
- W2145756561 hasRelatedWork W2054803354 @default.
- W2145756561 hasRelatedWork W2058212189 @default.
- W2145756561 hasRelatedWork W2098432798 @default.
- W2145756561 hasRelatedWork W2107726111 @default.
- W2145756561 hasRelatedWork W2119567691 @default.
- W2145756561 hasRelatedWork W2121863487 @default.
- W2145756561 hasRelatedWork W2125074935 @default.
- W2145756561 hasRelatedWork W2128281152 @default.
- W2145756561 hasRelatedWork W2160284799 @default.
- W2145756561 hasRelatedWork W2341171179 @default.
- W2145756561 hasRelatedWork W2884173129 @default.
- W2145756561 hasRelatedWork W3011120880 @default.
- W2145756561 hasRelatedWork W2189252197 @default.
- W2145756561 hasVolume "9" @default.
- W2145756561 isParatext "false" @default.
- W2145756561 isRetracted "false" @default.
- W2145756561 magId "2145756561" @default.
- W2145756561 workType "article" @default.