Data Reuse in Non-Stationary Learning
This paper proposes a class of anytime algorithms, dubbed Exposure-Capped Reuse (ECR), that combine online change detection, compatibility testing, and contamination control, and derives a novel information-theoretic lower bound that establishes the near-minimax optimality of ECR.
Tomer Gafni, Garud Iyengar, A. Zeevi
· 0 citations