<i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection
Vertical Federated Learning (VFL) is a promising category of Federated Learning that enables collaborative model training among distributed parties with data privacy protection. Due to its unique training architecture, a key challenge of VFL is high communication cost due to transmitting intermediat...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/27/1/66 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588595272089600 |
|---|---|
| author | Jiahui Zhou Han Liang Tian Wu Xiaoxi Zhang Yu Jiang Chee Wei Tan |
| author_facet | Jiahui Zhou Han Liang Tian Wu Xiaoxi Zhang Yu Jiang Chee Wei Tan |
| author_sort | Jiahui Zhou |
| collection | DOAJ |
| description | Vertical Federated Learning (VFL) is a promising category of Federated Learning that enables collaborative model training among distributed parties with data privacy protection. Due to its unique training architecture, a key challenge of VFL is high communication cost due to transmitting intermediate results between the Active Party and Passive Parties. Current communication-efficient VFL methods rely on using stale results without meticulous selection, which can impair model accuracy, particularly in noisy data environments. To address these limitations, this work proposes <i>VFL-Cafe</i>, a new VFL training method that leverages dynamic caching and feature selection to boost communication efficiency and model accuracy. In each communication round, the employed caching scheme allows multiple batches of intermediate results to be cached and strategically reused by different parties, reducing the communication overhead while maintaining model accuracy. Additionally, to eliminate the negative impact of noisy features that may undermine the effectiveness of using stale results to reduce communication rounds and incur significant model degradation, a feature selection strategy is integrated into each round of local updates. Theoretical analysis is then conducted to provide guidance on cache configuration, optimizing performance. Finally, extensive experimental results validate <i>VFL-Cafe</i>’s efficacy, demonstrating remarkable improvements in communication efficiency and model accuracy. |
| format | Article |
| id | doaj-art-f96d0191d58f44e08cd2d0934b6ca7f5 |
| institution | Kabale University |
| issn | 1099-4300 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj-art-f96d0191d58f44e08cd2d0934b6ca7f52025-01-24T13:31:52ZengMDPI AGEntropy1099-43002025-01-012716610.3390/e27010066<i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature SelectionJiahui Zhou0Han Liang1Tian Wu2Xiaoxi Zhang3Yu Jiang4Chee Wei Tan5School of Computer and Science and Engineering, Sun Yat-sen University, Guangzhou 510275, ChinaSchool of Computer and Science and Engineering, Sun Yat-sen University, Guangzhou 510275, ChinaSchool of Computer and Science and Engineering, Sun Yat-sen University, Guangzhou 510275, ChinaSchool of Computer and Science and Engineering, Sun Yat-sen University, Guangzhou 510275, ChinaCollege of Computing and Data Science, Nanyang Technological University in Singapore, Singapore 639798, SingaporeCollege of Computing and Data Science, Nanyang Technological University in Singapore, Singapore 639798, SingaporeVertical Federated Learning (VFL) is a promising category of Federated Learning that enables collaborative model training among distributed parties with data privacy protection. Due to its unique training architecture, a key challenge of VFL is high communication cost due to transmitting intermediate results between the Active Party and Passive Parties. Current communication-efficient VFL methods rely on using stale results without meticulous selection, which can impair model accuracy, particularly in noisy data environments. To address these limitations, this work proposes <i>VFL-Cafe</i>, a new VFL training method that leverages dynamic caching and feature selection to boost communication efficiency and model accuracy. In each communication round, the employed caching scheme allows multiple batches of intermediate results to be cached and strategically reused by different parties, reducing the communication overhead while maintaining model accuracy. Additionally, to eliminate the negative impact of noisy features that may undermine the effectiveness of using stale results to reduce communication rounds and incur significant model degradation, a feature selection strategy is integrated into each round of local updates. Theoretical analysis is then conducted to provide guidance on cache configuration, optimizing performance. Finally, extensive experimental results validate <i>VFL-Cafe</i>’s efficacy, demonstrating remarkable improvements in communication efficiency and model accuracy.https://www.mdpi.com/1099-4300/27/1/66vertical federated learningcommunication efficientfeature selectiondynamic caching |
| spellingShingle | Jiahui Zhou Han Liang Tian Wu Xiaoxi Zhang Yu Jiang Chee Wei Tan <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection Entropy vertical federated learning communication efficient feature selection dynamic caching |
| title | <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection |
| title_full | <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection |
| title_fullStr | <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection |
| title_full_unstemmed | <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection |
| title_short | <i>VFL-Cafe</i>: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection |
| title_sort | i vfl cafe i communication efficient vertical federated learning via dynamic caching and feature selection |
| topic | vertical federated learning communication efficient feature selection dynamic caching |
| url | https://www.mdpi.com/1099-4300/27/1/66 |
| work_keys_str_mv | AT jiahuizhou ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection AT hanliang ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection AT tianwu ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection AT xiaoxizhang ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection AT yujiang ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection AT cheeweitan ivflcafeicommunicationefficientverticalfederatedlearningviadynamiccachingandfeatureselection |