FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2987293847> ?p ?o ?g. }

• W2987293847 abstract "We consider the applications of the Frank-Wolfe (FW) algorithm for Apprenticeship Learning (AL). In this setting, we are given a Markov Decision Process (MDP) without an explicit reward function. Instead, we observe an expert that acts according to some policy, and the goal is to find a policy whose feature expectations are closest to those of the expert policy. We formulate this problem as finding the projection of the feature expectations of the expert on the feature expectations polytope -- the convex hull of the feature expectations of all the deterministic policies in the MDP. We show that this formulation is equivalent to the AL objective and that solving this problem using the FW algorithm is equivalent well-known Projection method of Abbeel and Ng (2004). This insight allows us to analyze AL with tools from convex optimization literature and derive tighter convergence bounds on AL. Specifically, we show that a variation of the FW method that is based on taking away steps achieves a linear rate of convergence when applied to AL and that a stochastic version of the FW algorithm can be used to avoid precise estimation of feature expectations. We also experimentally show that this version outperforms the FW baseline. To the best of our knowledge, this is the first work that shows linear convergence rates for AL." @default.
• W2987293847 created "2019-11-22" @default.
• W2987293847 creator A5006699796 @default.
• W2987293847 creator A5014637159 @default.
• W2987293847 creator A5018613019 @default.
• W2987293847 creator A5053884531 @default.
• W2987293847 date "2019-11-05" @default.
• W2987293847 modified "2023-09-27" @default.
• W2987293847 title "Apprenticeship Learning via Frank-Wolfe" @default.
• W2987293847 cites W1495909821 @default.
• W2987293847 cites W1505731132 @default.
• W2987293847 cites W1575592356 @default.
• W2987293847 cites W1969244384 @default.
• W2987293847 cites W1984901446 @default.
• W2987293847 cites W1999874108 @default.
• W2987293847 cites W2013243898 @default.
• W2987293847 cites W2028355641 @default.
• W2987293847 cites W2061562262 @default.
• W2987293847 cites W2078735805 @default.
• W2987293847 cites W2102847492 @default.
• W2987293847 cites W2113023245 @default.
• W2987293847 cites W2119567691 @default.
• W2987293847 cites W2134882453 @default.
• W2987293847 cites W2136885855 @default.
• W2987293847 cites W2141076336 @default.
• W2987293847 cites W2169498096 @default.
• W2987293847 cites W2187299368 @default.
• W2987293847 cites W2257807852 @default.
• W2987293847 cites W2481906283 @default.
• W2987293847 cites W2945884844 @default.
• W2987293847 cites W2962906893 @default.
• W2987293847 cites W905619 @default.
• W2987293847 hasPublicationYear "2019" @default.
• W2987293847 type Work @default.
• W2987293847 sameAs 2987293847 @default.
• W2987293847 citedByCount "0" @default.
• W2987293847 crossrefType "posted-content" @default.
• W2987293847 hasAuthorship W2987293847A5006699796 @default.
• W2987293847 hasAuthorship W2987293847A5014637159 @default.
• W2987293847 hasAuthorship W2987293847A5018613019 @default.
• W2987293847 hasAuthorship W2987293847A5053884531 @default.
• W2987293847 hasConcept C105795698 @default.
• W2987293847 hasConcept C106189395 @default.
• W2987293847 hasConcept C112680207 @default.
• W2987293847 hasConcept C11413529 @default.
• W2987293847 hasConcept C118615104 @default.
• W2987293847 hasConcept C126255220 @default.
• W2987293847 hasConcept C138885662 @default.
• W2987293847 hasConcept C145691206 @default.
• W2987293847 hasConcept C154945302 @default.
• W2987293847 hasConcept C159886148 @default.
• W2987293847 hasConcept C162324750 @default.
• W2987293847 hasConcept C206194317 @default.
• W2987293847 hasConcept C2524010 @default.
• W2987293847 hasConcept C26517878 @default.
• W2987293847 hasConcept C2776401178 @default.
• W2987293847 hasConcept C2777303404 @default.
• W2987293847 hasConcept C33923547 @default.
• W2987293847 hasConcept C38652104 @default.
• W2987293847 hasConcept C41008148 @default.
• W2987293847 hasConcept C41895202 @default.
• W2987293847 hasConcept C50522688 @default.
• W2987293847 hasConcept C57493831 @default.
• W2987293847 hasConcept C57869625 @default.
• W2987293847 hasConceptScore W2987293847C105795698 @default.
• W2987293847 hasConceptScore W2987293847C106189395 @default.
• W2987293847 hasConceptScore W2987293847C112680207 @default.
• W2987293847 hasConceptScore W2987293847C11413529 @default.
• W2987293847 hasConceptScore W2987293847C118615104 @default.
• W2987293847 hasConceptScore W2987293847C126255220 @default.
• W2987293847 hasConceptScore W2987293847C138885662 @default.
• W2987293847 hasConceptScore W2987293847C145691206 @default.
• W2987293847 hasConceptScore W2987293847C154945302 @default.
• W2987293847 hasConceptScore W2987293847C159886148 @default.
• W2987293847 hasConceptScore W2987293847C162324750 @default.
• W2987293847 hasConceptScore W2987293847C206194317 @default.
• W2987293847 hasConceptScore W2987293847C2524010 @default.
• W2987293847 hasConceptScore W2987293847C26517878 @default.
• W2987293847 hasConceptScore W2987293847C2776401178 @default.
• W2987293847 hasConceptScore W2987293847C2777303404 @default.
• W2987293847 hasConceptScore W2987293847C33923547 @default.
• W2987293847 hasConceptScore W2987293847C38652104 @default.
• W2987293847 hasConceptScore W2987293847C41008148 @default.
• W2987293847 hasConceptScore W2987293847C41895202 @default.
• W2987293847 hasConceptScore W2987293847C50522688 @default.
• W2987293847 hasConceptScore W2987293847C57493831 @default.
• W2987293847 hasConceptScore W2987293847C57869625 @default.
• W2987293847 hasLocation W29872938471 @default.
• W2987293847 hasOpenAccess W2987293847 @default.
• W2987293847 hasPrimaryLocation W29872938471 @default.
• W2987293847 hasRelatedWork W1516618490 @default.
• W2987293847 hasRelatedWork W2181692883 @default.
• W2987293847 hasRelatedWork W2401533533 @default.
• W2987293847 hasRelatedWork W2460675832 @default.
• W2987293847 hasRelatedWork W2491144192 @default.
• W2987293847 hasRelatedWork W2803448454 @default.
• W2987293847 hasRelatedWork W2947782105 @default.
• W2987293847 hasRelatedWork W2951992264 @default.
• W2987293847 hasRelatedWork W2979932837 @default.
• W2987293847 hasRelatedWork W2998111914 @default.

• next