Gradient Perturbation is Underrated for Differentially Private Convex Optimization

Open in new window