作者: Ohad Shamir, Liran Szlak
摘要
We propose an Online Learning with Local Permutations (OLLP) setting, in which the learner is allowed to slightly permute the