Branching bandit processes

Research output: Contribution to journalArticlepeer-review

Abstract

A set of niarms of type i, i = 1,…, L, is available. A pull of arm of type i occupies a duration Viat the end of which a reward Ci and Ni1,…, NiLnew arms are obtained, while all other arms are frozen. A Gittins priority order of types is obtained and shown to yield the maximal discounted reward from this branching process of arms.

Original languageEnglish
Pages (from-to)269-278
Number of pages10
JournalProbability in the Engineering and Informational Sciences
Volume2
Issue number3
DOIs
StatePublished - Jul 1988
Externally publishedYes

Bibliographical note

Funding Information:
This research has been partially supported by the National Science Foundation Grant ECS-8712798. © 1988 Cambridge University Press 0269-9648/88 $5.00 + .00 26 9

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Management Science and Operations Research
  • Industrial and Manufacturing Engineering

Fingerprint

Dive into the research topics of 'Branching bandit processes'. Together they form a unique fingerprint.

Cite this