Conceptpseudo-regret
Pseudo-Regret
A core performance metric for online learning algorithms, defined as the expected difference between the algorithm's cumulative loss and the cumulative loss of the best policy known in hindsight. It measures the gap between an algorithm's performance and that of the optimal strategy.
Timeline (last 90 days)
Oct 7
Under assumptions including bounded residuals and divergence, a fixed feasible model with optimal responses, free-query risk, and a linear model of optimal query risk, ORUCB can achieve a high-probability pseudo-regret bound of O(√T·log(T+1)) over T rounds.
Unverified50%