Projection Free Dynamic Online Learning

Projection-based algorithms are popular in the literature for online convex optimization with convex constraints which results in a bottleneck for the practical implementation of the algorithms. To avoid this bottleneck, we propose a projection-free scheme based on Frank-Wolfe: instead of online gradient steps, we use steps that are collinear with the gradient but guaranteed to be feasible. We establish performance in terms of dynamic regret, which quantifies cost accumulation as compared with the optimal at each individual time slot. Specifically, for convex losses, we establish $\ccalO(T^{1/2})$ dynamic regret up to metrics of non-stationarity. We relax the algorithm's required information to only noisy gradient estimates, i.e., partial feedback and derived the dynamic regret bounds. Experiments on matrix completion problem and background separation in the video demonstrate the favorable performance of the proposed scheme.