Assessing the Impacts of Imperfect Datasets on Client Selections in Federated Learning
2026-08-03 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors studied how differences and errors in the data from different clients affect the accuracy and speed of federated learning, a method where many users train models together without sharing their raw data. They found that picking which clients participate can help, but doing it unfairly can actually hurt performance. To fix this, they created a way to privately score each client's usefulness to improve the overall learning. Their experiments showed this scoring method works well in balancing fairness and accuracy.
federated learningnon-IID dataclient selectionmodel accuracyconvergencenoisy dataprivacy-preservingdata skewdistributed learning
Authors
Yuan-Heng Tsai, Li-Hsing Yen, Yan-Wei Chen
Abstract
Federated learning (FL) is a popular distributed learning framework where multiple clients perform local training and a server aggregates the locally updated models. FL enables decentralized training while preserving the privacy of clients' datasets. However, non-independent and identically distributed (non-IID) or noisy datasets can lead to low model accuracy or high convergence latency. Precluding these clients through client selection may mitigate the problem, but heavily biased client selections may also degrade the learning performance. In this study, we first experimentally measure the impact of non-IID data (including skews in data quantity and label distribution), noisy data, and fairness in client selection on model accuracy and convergence. We then propose a privacy-preserving scoring method to assess each client's contribution in FL, with experiments conducted to demonstrate the effectiveness of the proposed assessment.