Deep Subspace Clustering Method Based on Federated Learning Framework

DOI

10.2991/978-94-6463-823-3_103How to use a DOI?

Keywords

Federated Learning; Deep Clustering; Subspace Clustering

Abstract

This study focuses on deep subspace clustering in a federated learning framework, aiming to address the challenge of clustering high-dimensional data with privacy protection in a distributed environment. Traditional deep clustering methods usually have difficulty dealing with complex data scenarios and require centralized data access, which conflicts with privacy constraints. To overcome this limitation, Federated Deep Embedding Subspace Clustering (FedDESC) is proposed to achieve high-quality image clustering in a distributed environment with strong privacy protection. The method uses deep autoencoders to extract low-dimensional feature embeddings and integrates subspace clustering into the federated learning framework. Specifically, each client learns local embeddings and subspace basis in parallel, calculates reconstruction, affinity, and constraint losses, and updates parameters; the central server aggregates these updates using the Federated Averaging Algorithm (FedAvg) to form a global model. Experiments on Fashion- Modified National Institute of Standards and Technology database (MNIST), Canadian Institute for Advanced Research (CIFAR)-10, and CIFAR-100 datasets show that FedDESC achieves comparable clustering accuracy and normalized mutual information to centralized methods. These results confirm that FedDESC effectively balances clustering performance and privacy, providing a scalable solution for distributed data analysis.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Bokai Guo
PY  - 2025
DA  - 2025/08/31
TI  - Deep Subspace Clustering Method Based on Federated Learning Framework
BT  - Proceedings of the 2025 3rd International Conference on Image, Algorithms, and Artificial Intelligence (ICIAAI 2025)
PB  - Atlantis Press
SP  - 1070
EP  - 1079
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-823-3_103
DO  - 10.2991/978-94-6463-823-3_103
ID  - Guo2025
ER  -