
Research Article
Addressing Class Imbalance in Federated Learning via Collaborative GAN-Based Up-Sampling
@INPROCEEDINGS{10.1007/978-3-031-27041-3_15, author={Can Zhang and Xuefeng Liu and Shaojie Tang and Jianwei Niu and Tao Ren and Quanquan Hu}, title={Addressing Class Imbalance in Federated Learning via Collaborative GAN-Based Up-Sampling}, proceedings={Wireless Internet. 15th EAI International Conference, WiCON 2022, Virtual Event, November 2022, Proceedings}, proceedings_a={WICON}, year={2023}, month={2}, keywords={Federated learning Class imbalance Collaborative up-sampling Generative adversarial networks}, doi={10.1007/978-3-031-27041-3_15} }
- Can Zhang
Xuefeng Liu
Shaojie Tang
Jianwei Niu
Tao Ren
Quanquan Hu
Year: 2023
Addressing Class Imbalance in Federated Learning via Collaborative GAN-Based Up-Sampling
WICON
Springer
DOI: 10.1007/978-3-031-27041-3_15
Abstract
Federated learning (FL) is an emerging learning framework that enables decentralized devices to collaboratively train a model without leaking their data to each other. One common problem in FL is class imbalance, in which either the distribution or quantity of the training data varies in different devices. In the presence of class imbalance, the performance of the final model can be negatively affected. A straightforward approach to address class imbalance is up-sampling, by which data of minority classes in each device are augmented independently. However, this up-sampling approach does not allow devices to help each other and therefore its effectiveness can be greatly compromised. In this paper, we propose FED-CGU, a collaborative GAN-based up-sampling strategy in FL. In FED-CGU, devices can help each other during up-sampling via collaboratively training a GAN model which augments data for each device. In addition, some advanced designs of FED-CGU are proposed, including dynamically determining the number of augmented data in each device and selecting complementary devices that can better help each other. We test FED-CGU with benchmark datasets including Fashion-MNIST and CIFAR-10. Experimental results demonstrate that FED-CGU outperforms the state-of-the-art algorithms.