Best Arm Identification in Graphical Bilinear Bandits
Geovani Rizk 1 2 Albert Thomas 2 Igor Colin 2 Rida Laraki 1 3 Yann Chevaleyre 1
Abstract agent (e.g., all the configuration parameters of the antennas),
We introduce a new graphical bilinear bandit prob- and receives an associated global noisy reward (e.g., the
lem where a learner (or a central entity) allocates signal quality over the whole network). The goal of the
arms to the nodes ...


雷达卡




京公网安备 11010802022788号







