Finite-Time 4-Expert Prediction Problem
Abstract
We explicitly solve the nonlinear PDE that is the continuous limit of dynamic programming of \emph{expert prediction problem} in finite horizon setting with experts. The \emph{expert prediction problem} is formulated as a zero sum game between a player and an adversary. By showing that the solution is , we are able to show that the strategies conjectured in arXiv:1409.3040G form an asymptotic Nash equilibrium. We also prove the "Finite vs Geometric regret" conjecture proposed in arXiv:1409.3040G for , and and show that this conjecture in fact follows from the conjecture that the comb strategies are optimal.
Keywords
Cite
@article{arxiv.1911.10936,
title = {Finite-Time 4-Expert Prediction Problem},
author = {Erhan Bayraktar and Ibrahim Ekren and Xin Zhang},
journal= {arXiv preprint arXiv:1911.10936},
year = {2019}
}
Comments
Keywords: machine learning, expert advice framework, asymptotic expansion, inverse Laplace transform, regret minimization, Jacobi-theta function