We show that, surprisingly, the notion of optimal finite-time regret is not a uniquely defined property in this context and that, in general, it is decoupled from the asymptotic rate. We discuss alternative choices and propose a notion of finite-time optimality that we argue is meaningful .

Moran (NeurIPS 2019) showed that uniform or highly sample-efficient pure-private learners can be time-efficiently compiled into online learners.

It requires a bounded domain.

We study the problem of online learning with primary and secondary losses.

Much of the work in online learning focuses on the study of sublinear upper bounds on the regret.

Now we can formally analyze the feasibility of AD-OGP .

Most existing algorithms in decentralized online learning are conducted in the synchronous setting.

Player and Nature, where at each stage both pick actions based on the contexts.

Given such a setup, a natural question to ask is how does one measure the performance of the learner? Classical online learning studies one such notion of performance known as regret.