AI RESEARCH
Fixed Budget is No Harder Than Fixed Confidence in Best-Arm Identification up to Logarithmic Factors
arXiv CS.AI
•
ArXi:2602.03972v3 Announce Type: replace-cross The best-arm identification (BAI) problem is one of the most fundamental problems in interactive machine learning, which has two flavors: the fixed-budget setting (FB) and the fixed-confidence setting (FC). For $K$-armed bandits with a unique best arm, the optimal sample complexities for both settings have been settled down, and they match up to logarithmic factors.