Adaptivity and Non-stationarity: Problem-dependent Dynamic Regret for Online Convex Optimization

Open in new window