Scaling Reinforcement Learning Techniques via Modularity
DOI: 10.1016/B978-1-55860-200-7.50076-3
关键词: Aliasing (computing) 、 Task (project management) 、 Reinforcement learning 、 Variable (computer science) 、 Computer science 、 Robot 、 Variation (game tree) 、 Expected utility hypothesis 、 Artificial intelligence 、 Modularity (networks) 、 State vector
摘要: … of Q-learning that allows the modular architecture to reduce the effects of perceptual aliasing on reward estimation. Q-learning … actions that achieve a state in GNN · is the likelihood ratio …
elsevier.com
sci-hub.se 参考文章(5)
Richard S. Sutton, Integrated architecture for learning, planning, and reacting based on approximating dynamic programming international conference on machine learning. pp. 216- 224 ,(1990) , 10.1016/B978-1-55860-141-3.50030-4
Sridhar Mahadevan, Jonathan Connell, Automatic programming of behavior-based robots using reinforcement learning national conference on artificial intelligence. pp. 768- 773 ,(1991)
Richard E. Korf, Planning as search: a quantitative approach Artificial Intelligence. ,vol. 33, pp. 65- 88 ,(1987) , 10.1016/0004-3702(87)90051-8
Dana Ballard, Steven Whitehead, Learning to Perceive and Act ,(1990)
C. J. C. H. Watkins, Learning from delayed rewards Ph. D thesis, Cambridge University Psychology Department. ,(1989)