A Reliable Vertical Federated Learning Framework for Traffic State Estimation with Data Selection and Incentive Mechanisms
Zijun Zhan, Yaxian Dong, Daniel Mawunyo Doe, Yuqing Hu, Shuai Li, Shaohua Cao, and Zhu Han · 2025
Read A Reliable Vertical Federated Learning Framework for Traffic State Estimation with Data Selection and Incentive Mechanisms by Zijun Zhan, Yaxian Dong, Daniel Mawunyo Doe, Yuqing Hu, Shuai Li, Shaohua Cao, and Zhu Han (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-01285.txt.
Description
Zijun Zhan and co-authors wrote this 2025 paper on estimating traffic flow without pooling private data. City authorities and mobility firms each hold part of the traffic picture, and neither can simply hand its data over. Vertical federated learning lets them train one model while the data stay apart, but partners vary in quality and some may slack off.
The framework answers with two parts. First, the city splits the road network into segments and scores the data of each firm for each segment, picking the best provider per segment. Second, a penalty-based incentive, framed as a supervision game, discourages lazy data provision. The design aims to keep the shared model reliable during training and daily use.
Catalog record
| CREATOR | Zijun Zhan, Yaxian Dong, Daniel Mawunyo Doe, Yuqing Hu, Shuai Li, Shaohua Cao, and Zhu Han |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-01285