A Unified Benchmark of Federated Learning with Kolmogorov-Arnold Networks for Medical Imaging
This work addresses privacy-preserving medical imaging in distributed healthcare, but it is incremental as it applies an existing method (KAN) to a new setting (FL).
This paper tackles the problem of improving federated learning for medical imaging by evaluating Kolmogorov-Arnold Networks (KAN) against traditional MLPs across six FL algorithms on a blood cell classification dataset, demonstrating that KAN achieves superior performance with simpler architectures.
Federated Learning (FL) enables model training across decentralized devices without sharing raw data, thereby preserving privacy in sensitive domains like healthcare. In this paper, we evaluate Kolmogorov-Arnold Networks (KAN) architectures against traditional MLP across six state-of-the-art FL algorithms on a blood cell classification dataset. Notably, our experiments demonstrate that KAN can effectively replace MLP in federated environments, achieving superior performance with simpler architectures. Furthermore, we analyze the impact of key hyperparameters-grid size and network architecture-on KAN performance under varying degrees of Non-IID data distribution. In addition, our ablation studies reveal that optimizing KAN width while maintaining minimal depth yields the best performance in federated settings. As a result, these findings establish KAN as a promising alternative for privacy-preserving medical imaging applications in distributed healthcare. To the best of our knowledge, this is the first comprehensive benchmark of KAN in FL settings for medical imaging task.