Uncertainty quantification by block bootstrap for differentially private stochastic gradient descent
This work addresses uncertainty quantification for differentially private SGD, a critical issue for practitioners in privacy-sensitive domains, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackles the problem of uncertainty quantification for differentially private stochastic gradient descent, which was previously infeasible due to privacy constraints, and proposes a novel block bootstrap method that is computationally tractable and does not require adjusting the privacy budget, demonstrating its validity and finite sample properties through simulations.
Stochastic Gradient Descent (SGD) is a widely used tool in machine learning. In the context of Differential Privacy (DP), SGD has been well studied in the last years in which the focus is mainly on convergence rates and privacy guarantees. While in the non private case, uncertainty quantification (UQ) for SGD by bootstrap has been addressed by several authors, these procedures cannot be transferred to differential privacy due to multiple queries to the private data. In this paper, we propose a novel block bootstrap for SGD under local differential privacy that is computationally tractable and does not require an adjustment of the privacy budget. The method can be easily implemented and is applicable to a broad class of estimation problems. We prove the validity of our approach and illustrate its finite sample properties by means of a simulation study. As a by-product, the new method also provides a simple alternative numerical tool for UQ for non-private SGD.