Efficient, Adaptive Near-Field Beam Training based on Linear Bandit | Junchi Liu et al. | ResearchPod