FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3202588411> ?p ?o ?g. }

• W3202588411 abstract "Various automatic curriculum learning (ACL) methods have been proposed to improve the sample efficiency and final performance of deep reinforcement learning (DRL). They are designed to control how a DRL agent collects data, which is inspired by how humans gradually adapt their learning processes to their capabilities. For example, ACL can be used for subgoal generation, reward shaping, environment generation, or initial state generation. However, prior work only considers curriculum learning following one of the aforementioned predefined paradigms. It is unclear which of these paradigms are complementary, and how the combination of them can be learned from interactions with the environment. Therefore, in this paper, we propose a unified automatic curriculum learning framework to create multi-objective but coherent curricula that are generated by a set of parametric curriculum modules. Each curriculum module is instantiated as a neural network and is responsible for generating a particular curriculum. In order to coordinate those potentially conflicting modules in unified parameter space, we propose a multi-task hyper-net learning framework that uses a single hyper-net to parameterize all those curriculum modules. In addition to existing hand-designed curricula paradigms, we further design a flexible memory mechanism to learn an abstract curriculum, which may otherwise be difficult to design manually. We evaluate our method on a series of robotic manipulation tasks and demonstrate its superiority over other state-of-the-art ACL methods in terms of sample efficiency and final performance." @default.
• W3202588411 created "2021-10-11" @default.
• W3202588411 creator A5019685128 @default.
• W3202588411 creator A5020949970 @default.
• W3202588411 creator A5021430632 @default.
• W3202588411 creator A5033684035 @default.
• W3202588411 creator A5046206794 @default.
• W3202588411 creator A5053034244 @default.
• W3202588411 date "2021-10-06" @default.
• W3202588411 modified "2023-09-23" @default.
• W3202588411 title "Learning Multi-Objective Curricula for Deep Reinforcement Learning" @default.
• W3202588411 cites W1777239053 @default.
• W3202588411 cites W1863227302 @default.
• W3202588411 cites W2007347635 @default.
• W3202588411 cites W2012036715 @default.
• W3202588411 cites W2064675550 @default.
• W3202588411 cites W2089217417 @default.
• W3202588411 cites W2097451239 @default.
• W3202588411 cites W2112707476 @default.
• W3202588411 cites W2121863487 @default.
• W3202588411 cites W2151834591 @default.
• W3202588411 cites W2157331557 @default.
• W3202588411 cites W2169743339 @default.
• W3202588411 cites W2174786457 @default.
• W3202588411 cites W2296073425 @default.
• W3202588411 cites W2414711238 @default.
• W3202588411 cites W2436711315 @default.
• W3202588411 cites W24477102 @default.
• W3202588411 cites W2553882142 @default.
• W3202588411 cites W2581240229 @default.
• W3202588411 cites W2594466397 @default.
• W3202588411 cites W2624871570 @default.
• W3202588411 cites W2736601468 @default.
• W3202588411 cites W2737215407 @default.
• W3202588411 cites W2741594138 @default.
• W3202588411 cites W2751973545 @default.
• W3202588411 cites W2775536965 @default.
• W3202588411 cites W2785342287 @default.
• W3202588411 cites W2890538051 @default.
• W3202588411 cites W2892230114 @default.
• W3202588411 cites W2902334345 @default.
• W3202588411 cites W2902347140 @default.
• W3202588411 cites W2904157920 @default.
• W3202588411 cites W2913340405 @default.
• W3202588411 cites W2946054758 @default.
• W3202588411 cites W2950527759 @default.
• W3202588411 cites W2951008357 @default.
• W3202588411 cites W2962743139 @default.
• W3202588411 cites W2962974944 @default.
• W3202588411 cites W2963199420 @default.
• W3202588411 cites W2963276097 @default.
• W3202588411 cites W2963293881 @default.
• W3202588411 cites W2963341924 @default.
• W3202588411 cites W2963403593 @default.
• W3202588411 cites W2963577640 @default.
• W3202588411 cites W2963677766 @default.
• W3202588411 cites W2963820385 @default.
• W3202588411 cites W2964118020 @default.
• W3202588411 cites W2964118342 @default.
• W3202588411 cites W2964291307 @default.
• W3202588411 cites W2964327384 @default.
• W3202588411 cites W2964335674 @default.
• W3202588411 cites W2968917487 @default.
• W3202588411 cites W2972758308 @default.
• W3202588411 cites W2978409868 @default.
• W3202588411 cites W2989321827 @default.
• W3202588411 cites W2991420892 @default.
• W3202588411 cites W2997359900 @default.
• W3202588411 cites W2997970896 @default.
• W3202588411 cites W3012445938 @default.
• W3202588411 cites W3012934742 @default.
• W3202588411 cites W3013821552 @default.
• W3202588411 cites W3036282537 @default.
• W3202588411 cites W3040707741 @default.
• W3202588411 cites W3091948595 @default.
• W3202588411 cites W3099920963 @default.
• W3202588411 cites W3104924044 @default.
• W3202588411 cites W3127856697 @default.
• W3202588411 cites W567721252 @default.
• W3202588411 hasPublicationYear "2021" @default.
• W3202588411 type Work @default.
• W3202588411 sameAs 3202588411 @default.
• W3202588411 citedByCount "0" @default.
• W3202588411 crossrefType "posted-content" @default.
• W3202588411 hasAuthorship W3202588411A5019685128 @default.
• W3202588411 hasAuthorship W3202588411A5020949970 @default.
• W3202588411 hasAuthorship W3202588411A5021430632 @default.
• W3202588411 hasAuthorship W3202588411A5033684035 @default.
• W3202588411 hasAuthorship W3202588411A5046206794 @default.
• W3202588411 hasAuthorship W3202588411A5053034244 @default.
• W3202588411 hasBestOaLocation W32025884111 @default.
• W3202588411 hasConcept C119857082 @default.
• W3202588411 hasConcept C127413603 @default.
• W3202588411 hasConcept C154945302 @default.
• W3202588411 hasConcept C15744967 @default.
• W3202588411 hasConcept C177264268 @default.
• W3202588411 hasConcept C19417346 @default.
• W3202588411 hasConcept C199360897 @default.
• W3202588411 hasConcept C201995342 @default.
• W3202588411 hasConcept C2780451532 @default.

• next