Enhancing Accuracy in Differentially Private Distributed Optimization Through Sensitivity Reduction

2025/05/12 by Bin Liu, Xie, Furan, Liu, Bing +2
Computer Science · #Cryptography and Data Security #FOS: Mathematics #Optimization and Control (math.OC) #Privacy-Preserving Technologies in Data #Stochastic Gradient Optimization Techniques

paper · pdf · doi:10.48550/arxiv.2505.07482

openalex publication_date 2025/05/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we investigate the problem of differentially private distributed optimization. Recognizing that lower sensitivity leads to higher accuracy, we analyze the key factors influencing the sensitivity of differentially private distributed algorithms. Building on these insights, we propose a novel differentially private distributed algorithm for undirected graphs that enhances optimization accuracy by reducing sensitivity. To ensure practical applicability, we derive an explicit closed-form expression for the noise parameter as a function of the privacy budget. Moreover, we rigorously prove that the proposed algorithm can achieve arbitrarily rigorous ε-differential privacy, establish its convergence in the mean square sense, and provide an upper bound on its optimization accuracy. Finally, extensive comparisons with various privacy-preserving methods validate the effectiveness of our algorithm.

Citations

• QuanCrypt-FL: Quantized Homomorphic Encryption with Pruning for Secure Federated Learning
• Cryptography-Based Privacy-Preserving Method for Distributed Optimization over Time-Varying Directed Graphs with Enhanced Efficiency
• Privacy in Multi-agent Systems
• A Variance-Reduced Aggregation Based Gradient Tracking method for Distributed Optimization over Directed Networks
• Echo of Neighbors: Privacy Amplification for Personalized Private Federated Learning with Shuffle Model
• Gradient-tracking Based Differentially Private Distributed Optimization with Enhanced Optimization Accuracy
• Adap DP-FL: Differentially Private Federated Learning with Adaptive Noise
• Gradient-tracking based Distributed Optimization with Guaranteed Optimality under Noisy Information Sharing
• Dynamics based Privacy Preservation in Decentralized Optimization
• Tailoring Gradient Methods for Differentially-Private Distributed Optimization
• FLAME: Differentially Private Federated Learning in the Shuffle Model
• Privacy-Preserving Distributed Processing: Metrics, Bounds, and Algorithms
• Deep Leakage from Gradients
• Push-Pull Gradient Methods for Distributed Optimization in Networks
• Distributed Stochastic Gradient Tracking Methods
• Distributed stochastic gradient tracking methods
• Differentially Private Distributed Optimization
• Fast Distributed Gradient Methods

Related

• Gradient-tracking Based Differentially Private Distributed Optimization with Enhanced Optimization Accuracy
• Differentially Private Distributed Optimization
• Distributed Least-Squares Optimization Solvers with Differential Privacy
• Tailoring Gradient Methods for Differentially-Private Distributed Optimization
• Guaranteed Privacy of Distributed Nonconvex Optimization via Mixed-Monotone Functional Perturbations