Optimal Online Learning using Potential Functions
Machine Learning
2022-12-08 v6
Abstract
We study a family of potential functions for online learning. We show that if the potential function has strictly positive derivatives of order 1-4 then the min-max optimal strategy for the adversary is Brownian motion. Using that fact we analyze different potential functions and show that the Normal-Hedge potential provides the tightest upper bounds on the cumulative regret of the top {\epsilon}-percentile.
Keywords
Cite
@article{arxiv.2106.10717,
title = {Optimal Online Learning using Potential Functions},
author = {Yoav Freund},
journal= {arXiv preprint arXiv:2106.10717},
year = {2022}
}