22
41

Impossible Tuning Made Possible: A New Expert Algorithm and Its Applications

Abstract

We resolve the long-standing "impossible tuning" issue for the classic expert problem and show that, it is in fact possible to achieve regret O((lnd)tt,i2)O\left(\sqrt{(\ln d)\sum_t \ell_{t,i}^2}\right) simultaneously for all expert ii in a TT-round dd-expert problem where t,i\ell_{t,i} is the loss for expert ii in round tt. Our algorithm is based on the Mirror Descent framework with a correction term and a weighted entropy regularizer. While natural, the algorithm has not been studied before and requires a careful analysis. We also generalize the bound to O((lnd)t(t,imt,i)2)O\left(\sqrt{(\ln d)\sum_t (\ell_{t,i}-m_{t,i})^2}\right) for any prediction vector mtm_t that the learner receives, and recover or improve many existing results by choosing different mtm_t. Furthermore, we use the same framework to create a master algorithm that combines a set of base algorithms and learns the best one with little overhead. The new guarantee of our master allows us to derive many new results for both the expert problem and more generally Online Linear Optimization.

View on arXiv
Comments on this paper