Uncertainty quantification by block bootstrap for differentially private stochastic gradient descent

Open in new window