1992 QLearning

From GM-RKB
Jump to navigation Jump to search

Subject Headings: Q-Learning.

Notes

Cited By

Quotes

Abstract

[math]\displaystyle{ \cal{Q} }[/math]-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 [math]\displaystyle{ \cal Q }[/math]-learning based on that outlined in Watkins (1989). We show that [[[math]\displaystyle{ \cal Q }[/math]-learning converge]]s 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 [math]\displaystyle{ \cal Q }[/math] values can be changed each iteration, rather than just one.

References

;

 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
1992 QLearningChristopher J. C. H. Watkins
Peter Dayan
Technical Note : \cal Q -Learning10.1007/BF009926981992