A Huber Loss Minimization Approach to Mean Estimation under User-level Differential Privacy

Open in new window