On nearly selfoptimizing strategies for multiarmed bandit problems with controlled arms
Volume 23 / 1996
Applicationes Mathematicae 23 (1996), 449-473
DOI: 10.4064/am-23-4-449-473
Abstract
Two kinds of strategies for a multiarmed Markov bandit problem with controlled arms are considered: a strategy with forcing and a strategy with randomization. The choice of arm and control function in both cases is based on the current value of the average cost per unit time functional. Some simulation results are also presented.
