Optimal Best Markovian Arm Identification with
Fixed Confidence
Vrettos Moulos
Department of Electrical Engineering and Computer Sciences
University of California Berkeley
vrettos@berkeley.edu
Abstract
We give a complete characterization of the sampling complexity of best Markovian
arm identification in one-parameter Markovian bandit models. We derive instance
specific nonasymptoti ...


雷达卡



京公网安备 11010802022788号







