Balanced coarse-to-fine federated learning for noisy heterogeneous clients
Abstract For heterogeneous federated learning, each client cannot ensure the reliability due to the uncertainty in data collection, where different types of noise are always introduced into heterogeneous clients. Current existing methods rely on the specific assumptions for the distribution of noise...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-01-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-024-01694-8 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823861497070092288 |
|---|---|
| author | Longfei Han Ying Zhai Yanan Jia Qiang Cai Haisheng Li Xiankai Huang |
| author_facet | Longfei Han Ying Zhai Yanan Jia Qiang Cai Haisheng Li Xiankai Huang |
| author_sort | Longfei Han |
| collection | DOAJ |
| description | Abstract For heterogeneous federated learning, each client cannot ensure the reliability due to the uncertainty in data collection, where different types of noise are always introduced into heterogeneous clients. Current existing methods rely on the specific assumptions for the distribution of noise data to select the clean samples or eliminate noisy samples. However, heterogeneous clients have different deep neural network structures, and these models have different sensitivity to various noise types, the fixed noise-detection based methods may not be effective for each client. To overcome these challenges, we propose a balanced coarse-to-fine federated learning method to solve noisy heterogeneous clients. By introducing the coarse-to-fine two-stage strategy, the client can adaptively eliminate the noisy data. Meanwhile, we proposed a balanced progressive learning framework, It leverages the self-paced learning to sort the training samples from simple to difficult, which can evenly construct the client model from simple to difficult paradigm. The experimental results show that the proposed method has higher accuracy and robustness in processing noisy data from heterogeneous clients, and it is suitable for both heterogeneous and homogeneous federated learning scenarios. The code is avaliable at https://github.com/drafly/bcffl. |
| format | Article |
| id | doaj-art-ba54d3e0cf9e4075b431f4c3558579a1 |
| institution | Kabale University |
| issn | 2199-4536 2198-6053 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj-art-ba54d3e0cf9e4075b431f4c3558579a12025-02-09T13:00:58ZengSpringerComplex & Intelligent Systems2199-45362198-60532025-01-0111211210.1007/s40747-024-01694-8Balanced coarse-to-fine federated learning for noisy heterogeneous clientsLongfei Han0Ying Zhai1Yanan Jia2Qiang Cai3Haisheng Li4Xiankai Huang5School of Computer and Artificial Intelligence, Beijing Technology and Business UniversitySchool of Computer and Artificial Intelligence, Beijing Technology and Business UniversitySchool of Computer and Artificial Intelligence, Beijing Technology and Business UniversitySchool of Computer and Artificial Intelligence, Beijing Technology and Business UniversitySchool of Computer and Artificial Intelligence, Beijing Technology and Business UniversitySchool of Computer and Artificial Intelligence, Beijing Technology and Business UniversityAbstract For heterogeneous federated learning, each client cannot ensure the reliability due to the uncertainty in data collection, where different types of noise are always introduced into heterogeneous clients. Current existing methods rely on the specific assumptions for the distribution of noise data to select the clean samples or eliminate noisy samples. However, heterogeneous clients have different deep neural network structures, and these models have different sensitivity to various noise types, the fixed noise-detection based methods may not be effective for each client. To overcome these challenges, we propose a balanced coarse-to-fine federated learning method to solve noisy heterogeneous clients. By introducing the coarse-to-fine two-stage strategy, the client can adaptively eliminate the noisy data. Meanwhile, we proposed a balanced progressive learning framework, It leverages the self-paced learning to sort the training samples from simple to difficult, which can evenly construct the client model from simple to difficult paradigm. The experimental results show that the proposed method has higher accuracy and robustness in processing noisy data from heterogeneous clients, and it is suitable for both heterogeneous and homogeneous federated learning scenarios. The code is avaliable at https://github.com/drafly/bcffl.https://doi.org/10.1007/s40747-024-01694-8Noisy dataHeterogeneous clientsSelf-paced learningRobust federated learning |
| spellingShingle | Longfei Han Ying Zhai Yanan Jia Qiang Cai Haisheng Li Xiankai Huang Balanced coarse-to-fine federated learning for noisy heterogeneous clients Complex & Intelligent Systems Noisy data Heterogeneous clients Self-paced learning Robust federated learning |
| title | Balanced coarse-to-fine federated learning for noisy heterogeneous clients |
| title_full | Balanced coarse-to-fine federated learning for noisy heterogeneous clients |
| title_fullStr | Balanced coarse-to-fine federated learning for noisy heterogeneous clients |
| title_full_unstemmed | Balanced coarse-to-fine federated learning for noisy heterogeneous clients |
| title_short | Balanced coarse-to-fine federated learning for noisy heterogeneous clients |
| title_sort | balanced coarse to fine federated learning for noisy heterogeneous clients |
| topic | Noisy data Heterogeneous clients Self-paced learning Robust federated learning |
| url | https://doi.org/10.1007/s40747-024-01694-8 |
| work_keys_str_mv | AT longfeihan balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients AT yingzhai balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients AT yananjia balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients AT qiangcai balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients AT haishengli balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients AT xiankaihuang balancedcoarsetofinefederatedlearningfornoisyheterogeneousclients |