Conceptbest-of-both-worlds / 两全其美
BOBW
An algorithm design objective in online learning requiring a single algorithm to be near-optimal in both stochastic and adversarial environments without prior knowledge of the environment type.
Timeline (last 90 days)
Nov 3
Prior work (Lee et al. 2021, Kong & Zhou 2023, Ito et al. 2023, Zhao et al. 2025) failed to provide a BOBW algorithm within a pure FTRL framework that achieves instance-dependent optimal regret in stochastic environments; IAFTRL solves this open problem.
Unverified50%