Securing Distributed Gradient Descent in High Dimensional Statistical Learning

Lili Su, Jiaming Xu

We consider unreliable distributed learning systems wherein the training data is kept confidential by external workers, and the learner has to interact closely with those workers t…