1 problem

Let $d$ be the dimension, let $\calA$ denote the action space, and let $\mathfrak R_T^\star(d;\calA)$ be the minimax expected regret after $T$ rounds in stochastic bandit convex op…