Regret bounds of a distributed saddle point algorithm | IEEE Conference Publication | IEEE Xplore