Zobrazeno 1 - 10
of 12
pro vyhledávání: '"Hado Van Hasselt"'
Autor:
Wojciech Marian Czarnecki, Matteo Hessel, Lasse Espeholt, Hubert Soyer, Hado van Hasselt, Simon Schmitt
Publikováno v:
AAAI
The reinforcement learning (RL) community has made great strides in designing algorithms capable of exceeding human performance on specific tasks. These algorithms are mostly trained one task at the time, each new task requiring to train a brand new
Autor:
Borsa, Diana, Dabney, Will, Shaul, Tom, Silver, David, Munos, Rémi, Hunt, Jonathan, Quan, John, Hado Van Hasselt, Hessel, Matteo, Mankowitz, Daniel, Zídek, Augustin, Barreto, Andre
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::6cf7fdb2a948a39e321b5a61b27a1c2a
Autor:
Hado van Hasselt
Publikováno v:
Adaptation, Learning, and Optimization ISBN: 9783642276446
Reinforcement Learning
Reinforcement Learning
Many traditional reinforcement-learning algorithms have been designed for problems with small finite state and action spaces. Learning in such discrete problems can been difficult, due to noise and delayed reinforcements. However, many real-world pro
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::7376c10785672e077292fa116f2c22d6
https://doi.org/10.1007/978-3-642-27645-3_7
https://doi.org/10.1007/978-3-642-27645-3_7
Publikováno v:
Proceedings of the 2011 IEEE Symposium On Adaptive Dynamic Programming And Reinforcement Learning, 91-96
STARTPAGE=91;ENDPAGE=96;TITLE=Proceedings of the 2011 IEEE Symposium On Adaptive Dynamic Programming And Reinforcement Learning
ADPRL
STARTPAGE=91;ENDPAGE=96;TITLE=Proceedings of the 2011 IEEE Symposium On Adaptive Dynamic Programming And Reinforcement Learning
ADPRL
We describe a new framework for applying reinforcement learning (RL) algorithms to solve classification tasks by letting an agent act on the inputs and learn value functions. This paper describes how classification problems can be modeled using class
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a4b0c0580d1fc2caccc6b1ce60cfa4e6
https://research.rug.nl/en/publications/73fb908f-f10d-4087-bb48-12294313aad9
https://research.rug.nl/en/publications/73fb908f-f10d-4087-bb48-12294313aad9
Publikováno v:
Journal of Machine Learning Research, 12, 2045-2094. Microtome Publishing
University of Groningen
Journal of Machine Learning Research, June, 12, 2045-2094
Journal of Machine Learning Research, 12(Jun), 2045-2094
Journal of Machine Learning Research, 12, 2045-2094
University of Groningen
Journal of Machine Learning Research, June, 12, 2045-2094
Journal of Machine Learning Research, 12(Jun), 2045-2094
Journal of Machine Learning Research, 12, 2045-2094
This article presents and evaluates best-match learning, a new approach to reinforcement learning that trades off the sample efficiency of model-based methods with the space efficiency of model-free methods. Best-match learning works by approximating
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::7f1843b3e6684827e53786b0a9b42ac3
https://dare.uva.nl/personal/pure/en/publications/exploiting-bestmatch-equations-for-efficient-reinforcement-learning(dfece862-a70f-4d0a-91eb-a0a9377ca1f4).html
https://dare.uva.nl/personal/pure/en/publications/exploiting-bestmatch-equations-for-efficient-reinforcement-learning(dfece862-a70f-4d0a-91eb-a0a9377ca1f4).html
Autor:
Marco A. Wiering, Hado van Hasselt
Publikováno v:
IJCNN
Real-world control problems are often modeled as Markov Decision Processes (MDPs) with discrete action spaces to facilitate the use of the many reinforcement learning algorithms that exist to find solutions for such MDPs. For many of these problems a
Publikováno v:
Lecture Notes in Computer Science ISBN: 9783642111976
AGS
AGS
Increasing complexity in serious games and the need to reuse and adapt games to different purposes and different user needs, requires distributed development approaches. The use of software agents has been advocated as a means to deal with the comple
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::0b23b92ece90a6da15a601d1fc824d80
https://doi.org/10.1007/978-3-642-11198-3_14
https://doi.org/10.1007/978-3-642-11198-3_14
Autor:
Hado van Hasselt, Marco A. Wiering
Publikováno v:
ADPRL: 2009 IEEE SYMPOSIUM ON ADAPTIVE DYNAMIC PROGRAMMING AND REINFORCEMENT LEARNING, 101-108
STARTPAGE=101;ENDPAGE=108;TITLE=ADPRL: 2009 IEEE SYMPOSIUM ON ADAPTIVE DYNAMIC PROGRAMMING AND REINFORCEMENT LEARNING
ADPRL
STARTPAGE=101;ENDPAGE=108;TITLE=ADPRL: 2009 IEEE SYMPOSIUM ON ADAPTIVE DYNAMIC PROGRAMMING AND REINFORCEMENT LEARNING
ADPRL
This paper describes several new online model-free reinforcement learning (RL) algorithms. We designed three new reinforcement algorithms, namely: QV2, QVMAX, and QV-MAX2, that are all based on the QV-learning algorithm, but in contrary to QV-learnin
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::cb20aced6a9299b041e6ebae4373eb42
https://research.rug.nl/en/publications/f46669c4-8de6-4efb-ad92-e262c8d1732c
https://research.rug.nl/en/publications/f46669c4-8de6-4efb-ad92-e262c8d1732c
Publikováno v:
2009 IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning, ADPRL 2009, 30 March-2 April 2009, Nashville, TN, USA, 177-184
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning, 177-184
STARTPAGE=177;ENDPAGE=184;TITLE=Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning
University of Groningen
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning: ADPRL
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning
ADPRL
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning, 177-184
STARTPAGE=177;ENDPAGE=184;TITLE=Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning
University of Groningen
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning: ADPRL
Proceedings of the IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning
ADPRL
This paper presents a theoretical and empirical analysis of Expected Sarsa, a variation on Sarsa, the classic onpolicy temporal-difference method for model-free reinforcement learning. Expected Sarsa exploits knowledge about stochasticity in the beha
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::13a5d0c2059407fd54b09a93fb7338a7
http://resolver.tudelft.nl/uuid:0e868206-5237-42d0-8bc0-3deb94a679ef
http://resolver.tudelft.nl/uuid:0e868206-5237-42d0-8bc0-3deb94a679ef