Matches in SemOpenAlex for { <https://semopenalex.org/work/W3166618091> ?p ?o ?g. }
- W3166618091 abstract "We consider the problem of teaching via demonstrations in sequential decision-making settings. In particular, we study how to design a personalized curriculum over demonstrations to speed up the learner's convergence. We provide a unified curriculum strategy for two popular learner models: Maximum Causal Entropy Inverse Reinforcement Learning (MaxEnt-IRL) and Cross-Entropy Behavioral Cloning (CrossEnt-BC). Our unified strategy induces a ranking over demonstrations based on a notion of difficulty scores computed w.r.t. the teacher's optimal policy and the learner's current policy. Compared to the state of the art, our strategy doesn't require access to the learner's internal dynamics and still enjoys similar convergence guarantees under mild technical conditions. Furthermore, we adapt our curriculum strategy to teach a learner using domain knowledge in the form of task-specific difficulty scores when the teacher's optimal policy is unknown. Experiments on a car driving simulator environment and shortest path problems in a grid-world environment demonstrate the effectiveness of our proposed curriculum strategy." @default.
- W3166618091 created "2021-06-22" @default.
- W3166618091 creator A5005193721 @default.
- W3166618091 creator A5027711113 @default.
- W3166618091 creator A5040194824 @default.
- W3166618091 creator A5090876613 @default.
- W3166618091 date "2021-06-08" @default.
- W3166618091 modified "2023-09-24" @default.
- W3166618091 title "Curriculum Design for Teaching via Demonstrations: Theory and Applications" @default.
- W3166618091 cites W1541165948 @default.
- W3166618091 cites W1581407678 @default.
- W3166618091 cites W1929981607 @default.
- W3166618091 cites W1977655452 @default.
- W3166618091 cites W1986014385 @default.
- W3166618091 cites W1999874108 @default.
- W3166618091 cites W2020764470 @default.
- W3166618091 cites W2051228319 @default.
- W3166618091 cites W2051944318 @default.
- W3166618091 cites W2061562262 @default.
- W3166618091 cites W2062427462 @default.
- W3166618091 cites W2097625105 @default.
- W3166618091 cites W2098774185 @default.
- W3166618091 cites W2121863487 @default.
- W3166618091 cites W2133068870 @default.
- W3166618091 cites W2148112459 @default.
- W3166618091 cites W2148886952 @default.
- W3166618091 cites W2151834591 @default.
- W3166618091 cites W2152166054 @default.
- W3166618091 cites W2156163138 @default.
- W3166618091 cites W2162009473 @default.
- W3166618091 cites W2169528473 @default.
- W3166618091 cites W2174803659 @default.
- W3166618091 cites W2296073425 @default.
- W3166618091 cites W2557026499 @default.
- W3166618091 cites W2581240229 @default.
- W3166618091 cites W2616877139 @default.
- W3166618091 cites W2621205314 @default.
- W3166618091 cites W2735318784 @default.
- W3166618091 cites W2751516180 @default.
- W3166618091 cites W2786676179 @default.
- W3166618091 cites W2787244454 @default.
- W3166618091 cites W2794908222 @default.
- W3166618091 cites W2803740478 @default.
- W3166618091 cites W2808642110 @default.
- W3166618091 cites W2890752237 @default.
- W3166618091 cites W2902334345 @default.
- W3166618091 cites W2912555327 @default.
- W3166618091 cites W2916746008 @default.
- W3166618091 cites W2917620156 @default.
- W3166618091 cites W2942337198 @default.
- W3166618091 cites W2945995507 @default.
- W3166618091 cites W2962687617 @default.
- W3166618091 cites W2962930238 @default.
- W3166618091 cites W2963081269 @default.
- W3166618091 cites W2963293881 @default.
- W3166618091 cites W2963308241 @default.
- W3166618091 cites W2963311874 @default.
- W3166618091 cites W2963577640 @default.
- W3166618091 cites W2963590100 @default.
- W3166618091 cites W2964118020 @default.
- W3166618091 cites W2964122237 @default.
- W3166618091 cites W2964327384 @default.
- W3166618091 cites W2966120739 @default.
- W3166618091 cites W2970469191 @default.
- W3166618091 cites W2970734210 @default.
- W3166618091 cites W3085438811 @default.
- W3166618091 cites W3116940672 @default.
- W3166618091 cites W3128970217 @default.
- W3166618091 cites W3159530976 @default.
- W3166618091 cites W3165687057 @default.
- W3166618091 cites W3170354064 @default.
- W3166618091 cites W64088143 @default.
- W3166618091 cites W950880443 @default.
- W3166618091 cites W2963654306 @default.
- W3166618091 hasPublicationYear "2021" @default.
- W3166618091 type Work @default.
- W3166618091 sameAs 3166618091 @default.
- W3166618091 citedByCount "0" @default.
- W3166618091 crossrefType "posted-content" @default.
- W3166618091 hasAuthorship W3166618091A5005193721 @default.
- W3166618091 hasAuthorship W3166618091A5027711113 @default.
- W3166618091 hasAuthorship W3166618091A5040194824 @default.
- W3166618091 hasAuthorship W3166618091A5090876613 @default.
- W3166618091 hasConcept C127413603 @default.
- W3166618091 hasConcept C145420912 @default.
- W3166618091 hasConcept C154945302 @default.
- W3166618091 hasConcept C15744967 @default.
- W3166618091 hasConcept C162324750 @default.
- W3166618091 hasConcept C189430467 @default.
- W3166618091 hasConcept C19417346 @default.
- W3166618091 hasConcept C201995342 @default.
- W3166618091 hasConcept C2777303404 @default.
- W3166618091 hasConcept C2780451532 @default.
- W3166618091 hasConcept C33923547 @default.
- W3166618091 hasConcept C41008148 @default.
- W3166618091 hasConcept C47177190 @default.
- W3166618091 hasConcept C50522688 @default.
- W3166618091 hasConcept C97541855 @default.
- W3166618091 hasConceptScore W3166618091C127413603 @default.
- W3166618091 hasConceptScore W3166618091C145420912 @default.