Year: 2022
Type: preprint
Abstract: Federated Learning (FL) is a distributed machine learning paradigm where clients collaboratively train a model using their local (human-generated) datasets. While existing studies focus on FL algorithm development to tackle data heterogeneity across clients, the important issue of data quality (e.g., label noise) in FL is overlooked. This paper aims to fill this gap by providing a quantitative study on the impact of label noise on FL. We derive an upper bound for the generalization error that is linear in the clients' label noise level. Then we conduct experiments on MNIST and CIFAR-10 datasets using various FL algorithms. Our empirical results show that the global model accuracy linearly decreases as the noise level increases, which is consistent with our theoretical analysis. We further find that label noise slows down the convergence of FL training, and the global model tends to overfit when the noise level is high.
Source: arXiv (Cornell University)
Authors: Ke, Shuqi, Huang, Chao, Liu, Xin
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: Track-D: Data-Driven Disease Prevention and Control in Animal Health, G1851784936, G2958765513, BIGDATA: IA: A multi-level approach for global optimization of the surveillance and control of infectious disease in the swine industry, CNS Core: Medium: Collaborative: Exploring and Exploiting Learning for Efficient Network Control: Non-Stationarity, Inter-Dependence, and Domain-Knowledge, G5620962805
OpenAlex ID: W4309201601