The paper discusses the development of a new federated clustering method, CCFC. The method combines federated clustering and representation learning to improve the clustering performance. CCFC has shown significant improvements over baseline methods in handling complex, high-dimensional data and dealing with device failures. The study highlights the necessity and challenges of combining federated clustering and representation learning, and illustrates these with simulated scenarios on MNIST.
Publication date: 12 Jan 2024
Project Page: https://arxiv.org/abs/2401.06634
Paper: https://arxiv.org/pdf/2401.06634
