Balancing Utility and Privacy: Dynamically Private SGD with Random Projection
This work addresses privacy leakage concerns in machine learning models for users needing secure optimization, though it appears incremental as it builds on existing differentially private SGD methods.
The paper tackles the trade-off between utility and privacy in stochastic gradient descent by introducing D2P2-SGD, which combines dynamic differential privacy with random projection to achieve provably sub-linear convergence rates and enhances accuracy across diverse datasets while maintaining privacy.
Stochastic optimization is a pivotal enabler in modern machine learning, producing effective models for various tasks. However, several existing works have shown that model parameters and gradient information are susceptible to privacy leakage. Although Differentially Private SGD (DPSGD) addresses privacy concerns, its static noise mechanism impacts the error bounds for model performance. Additionally, with the exponential increase in model parameters, efficient learning of these models using stochastic optimizers has become more challenging. To address these concerns, we introduce the Dynamically Differentially Private Projected SGD (D2P2-SGD) optimizer. In D2P2-SGD, we combine two important ideas: (i) dynamic differential privacy (DDP) with automatic gradient clipping and (ii) random projection with SGD, allowing dynamic adjustment of the tradeoff between utility and privacy of the model. It exhibits provably sub-linear convergence rates across different objective functions, matching the best available rate. The theoretical analysis further suggests that DDP leads to better utility at the cost of privacy, while random projection enables more efficient model learning. Extensive experiments across diverse datasets show that D2P2-SGD remarkably enhances accuracy while maintaining privacy. Our code is available here.