logarithmic regret bound

position-based model (1) unimodal bandit (1) online ranking (1) click-through rate (1) bandit algorithm (1) regret bound (1) multi-armed bandit (1) learning to rank (1) click modeling (1) online learning (1)