Tight Lower Bounds for Multiplicative Weights Algorithmic Families

We study the fundamental problem of prediction with expert advice and develop regret lower bounds for a large family of algorithms for this problem. We develop simple adversarial primitives, that lend themselves to various combinations leading to sharp lower bounds for many algorithmic families. We use these primitives to show that the classic Multiplicative Weights Algorithm (MWA) has a regret of √ T ln k 2 , there by completely closing the gap between upper and lower bounds. We further show a regret lower bound of 2 3 √ T ln k 2 for a much more general family of algorithms than MWA, where the learning rate can be arbitrarily varied over time, or even picked from arbitrary distributions over time. We also use our primitives to construct adversaries in the geometric horizon setting for MWA to precisely characterize the regret at 0.391 √ δ for the case of 2 experts and a lower bound of 12 √ ln k 2δ for the case of arbitrary number of experts k. ∗Massachusetts Institute of Technology. 32 Vassar St, Cambridge, MA 02139. [email protected]. †Microsoft Research. One Microsoft Way, Redmond, WA 98052. [email protected]. ‡Google Research. 111 8th Avenue, New York, NY 10011. [email protected]. ar X iv :1 60 7. 02 83 4v 1 [ cs .L G ] 1 1 Ju l 2 01 6