305
v1v2 (latest)

Unconstrained Online Linear Learning in Hilbert Spaces: Minimax Algorithms and Normal Approximations

Annual Conference Computational Learning Theory (COLT), 2014
Abstract

We study algorithms for online linear optimization in Hilbert spaces, focusing on the case where the player is unconstrained. We develop a novel characterization of a large class of minimax algorithms, recovering, and even improving, several previous results as immediate corollaries. Moreover, using our tools, we develop an algorithm that provides a regret bound of O(UTlog(UTlog2T+1))\mathcal{O}\Big(U \sqrt{T \log(U \sqrt{T} \log^2 T +1)}\Big), where UU is the L2L_2 norm of an arbitrary comparator and both TT and UU are unknown to the player. This bound is optimal up to loglogT\sqrt{\log \log T} terms. When TT is known, we derive an algorithm with an optimal regret bound (up to constant factors). For both the known and unknown TT case, a Normal approximation to the conditional value of the game proves to be the key analysis tool.

View on arXiv
Comments on this paper