On the identification and mitigation of weaknesses in the Knowledge Gradient policy for multi-armed bandits

Edwards, James and Fearnhead, Paul and Glazebrook, Kevin David (2017) On the identification and mitigation of weaknesses in the Knowledge Gradient policy for multi-armed bandits. Probability in the Engineering and Informational Sciences, 31 (2). pp. 239-263. ISSN 0269-9648

[img]
Preview
PDF (1607.05970v1)
1607.05970v1.pdf - Accepted Version
Available under License Creative Commons Attribution-NonCommercial.

Download (468kB)

Abstract

The Knowledge Gradient (KG) policy was originally proposed for online ranking and selection problems but has recently been adapted for use in online decision making in general and multi-armed bandit problems (MABs) in particular. We study its use in a class of exponential family MABs and identify weaknesses, including a propensity to take actions which are dominated with respect to both exploitation and exploration. We propose variants of KG which avoid such errors. These new policies include an index heuristic which deploys a KG approach to develop an approximation to the Gittins index. A numerical study shows this policy to perform well over a range of MABs including those for which index policies are not optimal. While KG does not make dominated actions when bandits are Gaussian, it fails to be index consistent and appears not to enjoy a performance advantage over competitor policies when arms are correlated to compensate for its greater computational demands.

Item Type:
Journal Article
Journal or Publication Title:
Probability in the Engineering and Informational Sciences
Additional Information:
https://www.cambridge.org/core/journals/probability-in-the-engineering-and-informational-sciences The final, definitive version of this article has been published in the Journal,Probability in the Engineering and Informational Sciences, 31 (2), pp 239-263 2017, © 2016 Cambridge University Press.
Uncontrolled Keywords:
/dk/atira/pure/subjectarea/asjc/1800/1804
Subjects:
ID Code:
80524
Deposited By:
Deposited On:
30 Aug 2016 10:04
Refereed?:
Yes
Published?:
Published
Last Modified:
03 Dec 2020 03:23