flâneur — a map of the web's best reading

Technical Note: Q-Learning | Machine Learning

link.springer.com · 201 words · saved by 1 readers

𝑄 -learning (Watkins, 1989) is a simple way for agents to learn how to act optimally in controlled Markovian domains. It amounts to an incremental method for dynamic programming which imposes limited computational demands. It works by successively improving its evaluations of the quality of particular actions at particular states. This paper presents and proves in detail a convergence theorem for 𝑄 -learning based on that outlined in Watkins (1989). We show that 𝑄 -learning converges to the optimum action-values with probability 1 so long as all actions are repeatedly sampled in all states and the action-values are represented discretely. We also sketch extensions to the cases of non-discounted, but absorbing, Markov environments, and where many 𝑄 values can be changed each iteration, rather than just one. Download to read the full article text Avoid common mistakes on your manuscript. Barto, A.G., Bradtke, S.J. & Singh, S.P. (1991). Real-time l

Technical Note: Q-Learning Published: May 1992 Volume 8 , pages 279–292 ( 1992 ) Cite this article Download PDF Save article View saved research Machine Learning Aims and scope Submit manuscript Technical Note: Q-Learning Download PDF Abstract \(\mathcal{Q}\) -learning (Watkins, 1989) is a simple way for agents to learn how to act optimally in controlled Markovian domains. It amounts to an incremental method for dynamic programming which imposes limited computational demands. It works by successively improving its evaluations of the quality of particular actions at particular states. This pape

Explore this link on the map →

related reading