ConceptExponential Weights Algorithm / EWA
Exponential Weights Algorithm
A classical algorithm in online learning that updates the probability of strategy selection through exponential weighting of historical losses. It achieves optimal regret rates on a variety of online learning problems; in this work its dimension dependence is proven to be already optimal.
Timeline (last 90 days)
Oct 16
The regret rates of the exponentiated weights algorithm and the algorithm of Tseng et al. (2025) are already optimal in their polynomial dependence on the dimension d, matching the lower bound.
Unverified50%