Multi-armed bandits in discrete and continuous time (Q1296724): Difference between revisions
From MaRDI portal
Latest revision as of 20:26, 28 May 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Multi-armed bandits in discrete and continuous time |
scientific article |
Statements
Multi-armed bandits in discrete and continuous time (English)
0 references
19 July 2000
0 references
This paper analyzes Gittin's Markovian model in discrete and continuous time. Using multi-parameter processes and excursion theory, the authors show that index-priority strategies are optimal and express the optimal value in terms of indices.
0 references
multi-armed bandits
0 references
optimal increasing paths
0 references
multiparameter process
0 references
excursion
0 references
local time
0 references
dual predictable projection
0 references