FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2955664501> ?p ?o ?g. }

• W2955664501 abstract "In this work we present a novel approach for transfer-guided exploration in reinforcement learning that is inspired by the human tendency to leverage experiences from similar encounters in the past while navigating a new task. Given an optimal policy in a related task-environment, we show that its bisimulation distance from the current task-environment gives a lower bound on the optimal advantage of state-action pairs in the current task-environment. Transfer-guided Exploration (ExTra) samples actions from a Softmax distribution over these lower bounds. In this way, actions with potentially higher optimum advantage are sampled more frequently. In our experiments on gridworld environments, we demonstrate that given access to an optimal policy in a related task-environment, ExTra can outperform popular domain-specific exploration strategies viz. epsilon greedy, Model-Based Interval Estimation - Exploration Bonus (MBIE-EB), Pursuit and Boltzmann in rate of convergence. We further show that ExTra is robust to choices of source task and shows a graceful degradation of performance as the dissimilarity of the source task increases. We also demonstrate that ExTra, when used alongside traditional exploration algorithms, improves their rate of convergence. Thus it is capable of complementing the efficacy of traditional exploration algorithms." @default.
• W2955664501 created "2019-07-12" @default.
• W2955664501 creator A5009374923 @default.
• W2955664501 creator A5012951358 @default.
• W2955664501 creator A5032809289 @default.
• W2955664501 creator A5065952190 @default.
• W2955664501 date "2019-06-27" @default.
• W2955664501 modified "2023-09-27" @default.
• W2955664501 title "ExTra: Transfer-guided Exploration" @default.
• W2955664501 cites W1487219635 @default.
• W2955664501 cites W1491843047 @default.
• W2955664501 cites W1504212531 @default.
• W2955664501 cites W1504915502 @default.
• W2955664501 cites W1534331386 @default.
• W2955664501 cites W1559169059 @default.
• W2955664501 cites W1850488217 @default.
• W2955664501 cites W1982898256 @default.
• W2955664501 cites W1988526405 @default.
• W2955664501 cites W2000514530 @default.
• W2955664501 cites W2015667537 @default.
• W2955664501 cites W2031727428 @default.
• W2955664501 cites W2032916024 @default.
• W2955664501 cites W2034806191 @default.
• W2955664501 cites W2052117683 @default.
• W2955664501 cites W2058735307 @default.
• W2955664501 cites W2072229133 @default.
• W2955664501 cites W2097381042 @default.
• W2955664501 cites W2099111195 @default.
• W2955664501 cites W2105960367 @default.
• W2955664501 cites W2106639887 @default.
• W2955664501 cites W2113953866 @default.
• W2955664501 cites W2121863487 @default.
• W2955664501 cites W2124352385 @default.
• W2955664501 cites W2164114810 @default.
• W2955664501 cites W24272225 @default.
• W2955664501 cites W2489939061 @default.
• W2955664501 cites W2539033431 @default.
• W2955664501 cites W2568646110 @default.
• W2955664501 cites W2574473771 @default.
• W2955664501 cites W2623491082 @default.
• W2955664501 cites W2724169821 @default.
• W2955664501 cites W2769648743 @default.
• W2955664501 cites W2887671224 @default.
• W2955664501 cites W2945322269 @default.
• W2955664501 cites W2947766526 @default.
• W2955664501 cites W2948210913 @default.
• W2955664501 cites W2962817122 @default.
• W2955664501 cites W2963099933 @default.
• W2955664501 cites W2963176272 @default.
• W2955664501 cites W2963938771 @default.
• W2955664501 cites W36691172 @default.
• W2955664501 cites W79394677 @default.
• W2955664501 cites W946252087 @default.
• W2955664501 hasPublicationYear "2019" @default.
• W2955664501 type Work @default.
• W2955664501 sameAs 2955664501 @default.
• W2955664501 citedByCount "1" @default.
• W2955664501 countsByYear W29556645012019 @default.
• W2955664501 crossrefType "posted-content" @default.
• W2955664501 hasAuthorship W2955664501A5009374923 @default.
• W2955664501 hasAuthorship W2955664501A5012951358 @default.
• W2955664501 hasAuthorship W2955664501A5032809289 @default.
• W2955664501 hasAuthorship W2955664501A5065952190 @default.
• W2955664501 hasConcept C108583219 @default.
• W2955664501 hasConcept C119857082 @default.
• W2955664501 hasConcept C126255220 @default.
• W2955664501 hasConcept C127162648 @default.
• W2955664501 hasConcept C127413603 @default.
• W2955664501 hasConcept C150899416 @default.
• W2955664501 hasConcept C153083717 @default.
• W2955664501 hasConcept C154945302 @default.
• W2955664501 hasConcept C162324750 @default.
• W2955664501 hasConcept C188441871 @default.
• W2955664501 hasConcept C201995342 @default.
• W2955664501 hasConcept C2777303404 @default.
• W2955664501 hasConcept C2780451532 @default.
• W2955664501 hasConcept C31258907 @default.
• W2955664501 hasConcept C33923547 @default.
• W2955664501 hasConcept C41008148 @default.
• W2955664501 hasConcept C50522688 @default.
• W2955664501 hasConcept C57869625 @default.
• W2955664501 hasConcept C80444323 @default.
• W2955664501 hasConcept C97541855 @default.
• W2955664501 hasConceptScore W2955664501C108583219 @default.
• W2955664501 hasConceptScore W2955664501C119857082 @default.
• W2955664501 hasConceptScore W2955664501C126255220 @default.
• W2955664501 hasConceptScore W2955664501C127162648 @default.
• W2955664501 hasConceptScore W2955664501C127413603 @default.
• W2955664501 hasConceptScore W2955664501C150899416 @default.
• W2955664501 hasConceptScore W2955664501C153083717 @default.
• W2955664501 hasConceptScore W2955664501C154945302 @default.
• W2955664501 hasConceptScore W2955664501C162324750 @default.
• W2955664501 hasConceptScore W2955664501C188441871 @default.
• W2955664501 hasConceptScore W2955664501C201995342 @default.
• W2955664501 hasConceptScore W2955664501C2777303404 @default.
• W2955664501 hasConceptScore W2955664501C2780451532 @default.
• W2955664501 hasConceptScore W2955664501C31258907 @default.
• W2955664501 hasConceptScore W2955664501C33923547 @default.
• W2955664501 hasConceptScore W2955664501C41008148 @default.
• W2955664501 hasConceptScore W2955664501C50522688 @default.

• next