paper-with-me

홈 › Papers

Adaptive Methods Are Preferable in High Privacy Settings: An SDE Perspective

2026-03-03 · Enea Monzio Compagnoni, Alessandro Stanghellini, Rustem Islamov, Aurelien Lucchi, Anastasiia Koloskova arxiv

Differential Privacy (DP) is becoming central to large-scale training as privacy regulations tighten. We revisit how DP noise interacts with adaptivity in optimization through the lens of stochastic differential equations, providing the first SDE-based analysis of private optimizers. Focusing on DP-SGD and DP-SignSGD under per-example clipping, we show a sharp contrast under fixed hyperparameters: DP-SGD converges at a Privacy-Utility Trade-Off of $\mathcal{O}(1/\varepsilon^2)$ with speed independent of $\varepsilon$, while DP-SignSGD converges at a speed linear in $\varepsilon$ with an $\mathcal{O}(1/\varepsilon)$ trade-off, dominating in high-privacy or large batch noise regimes. By contrast, under optimal learning rates, both methods achieve comparable theoretical asymptotic performance; however, the optimal learning rate of DP-SGD scales linearly with $\varepsilon$, while that of DP-SignSGD is essentially $\varepsilon$-independent. This makes adaptive methods far more practical, as their hyperparameters transfer across privacy levels with little or no re-tuning. Empirical results confirm our theory across training and test metrics, and empirically extend from DP-SignSGD to DP-Adam.

📄 PDF Abstract BibTeX arXiv:2603.03226

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Federated Active Learning Under Extreme Non-IID and Global Class Imbalance

2026-03-11 · Chen-Chen Zong, Sheng-Jun Huang arxiv

Federated active learning (FAL) seeks to reduce annotation cost under privacy constraints, yet its effectiveness degrades in realistic settings with severe global class imbalance and highly heterogeneous clients. We cond…

Active Learning

Private Adaptive Optimization with Side Information

2022-02-12 · Tian Li, Manzil Zaheer, Sashank J. Reddi, Virginia Smith

Adaptive optimization methods have become the default solvers for many machine learning tasks. Unfortunately, the benefits of adaptivity may degrade when training with differential privacy, as the noise added to ensure p…

Combining Public and Private Data

2021-10-29 · Cecilia Ferrando, Jennifer Gillenwater, Alex Kulesza

Differential privacy is widely adopted to provide provable privacy guarantees in data analysis. We consider the problem of combining public and private data (and, more generally, data with heterogeneous privacy needs) fo…

Learning adaptive differential evolution algorithm from optimization experiences by policy gradient

2021-02-06 · Jianyong Sun, Xin Liu, Thomas Bäck, Zongben Xu

Differential evolution is one of the most prestigious population-based stochastic optimization algorithm for black-box problems. The performance of a differential evolution algorithm depends highly on its mutation and cr…

Evolutionary AlgorithmsStochastic Optimization

Balancing Privacy-Quality-Efficiency in Federated Learning through Round-Based Interleaving of Protection Techniques

2026-03-05 · Yenan Wang, Carla Fabiana Chiasserini, Elad Michael Schiller arxiv

In federated learning (FL), balancing privacy protection, learning quality, and efficiency remains a challenge. Privacy protection mechanisms, such as Differential Privacy (DP), degrade learning quality, or, as in the ca…

Federated Learning