LIA: Privacy-Preserving Data Quality Evaluation in Federated Learning Using a Lazy Influence Approximation

Abstract

In Federated Learning, it is crucial to handle low-quality, corrupted, or malicious data. However, traditional data valuation methods are not suitable due to privacy concerns. To address this, we propose a simple yet effective approach that utilizes a new influence approximation called "lazy influence" to filter and score data while preserving privacy. To do this, each participant uses their own data to estimate the influence of another participant's batch and sends a differentially private obfuscated score to the central coordinator. Our method has been shown to successfully filter out biased and corrupted data in various simulated and real-world settings, achieving a recall rate of over $>90\%$ (sometimes up to $100\%$) while maintaining strong differential privacy guarantees with $\epsilon \le 1$.

Proceedings of the 2024 IEEE International Conference on Big Data (IEEE BigData 2024). A preliminary version of this work received the Best Paper Award at the International Workshop on Trustworthy Federated Learning at IJCAI (FL-IJCAI) 2023

References in corpus (12)

• RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response
• Privacy Loss in Apple's Implementation of Differential Privacy on MacOS 10.12
• A Field Guide to Federated Optimization
• Federated Learning Of Out-Of-Vocabulary Words
• Unlocking High-Accuracy Differentially Private Image Classification through Scale
• On Privacy and Personalization in Cross-Silo Federated Learning
• Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning
• DMLR: Data-centric Machine Learning Research -- Past, Present and Future
• TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels
• LIA: Privacy-Preserving Data Quality Evaluation in Federated Learning Using a Lazy Influence Approximation
• Differentially Private Shapley Values for Data Evaluation
• Evaluating and Rewarding Teamwork Using Cooperative Game Abstractions

Cited by in corpus (2)

• LIA: Privacy-Preserving Data Quality Evaluation in Federated Learning Using a Lazy Influence Approximation
• Lazy But Effective: Collaborative Personalized Federated Learning with Heterogeneous Data