Tag
This paper studies multi-agent multi-armed bandits with heavy-tailed rewards under three information-asymmetry regimes, proposing robust decentralized algorithms with regret guarantees nearly matching centralized rates, and validating them on Pareto-distributed reward environments.
This paper studies cooperative multi-player bandits in continuous Lipschitz action spaces when the Lipschitz constant is unknown, proposing a meta-algorithm (mECAB) that estimates the constant and coordinates discretization across players under different information structures, with regret guarantees.