Federated Learning for Traffic Flow Prediction with Synthetic Data Augmentation
Fermin Orozco, Hongkai Wen, Pedro Porto Buarque de Gusm̃ao, Johan Wahlström, Man Luo · 2024
Read Federated Learning for Traffic Flow Prediction with Synthetic Data Augmentation by Fermin Orozco, Hongkai Wen, Pedro Porto Buarque de Gusm̃ao, Johan Wahlström, Man Luo (2024) 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-2412-08460.txt.
Description
Fermin Orozco and co-authors predict traffic flow with federated learning plus made up data in 2024, using team framing since the author order in the head differs from the catalog list. Traffic data is rich but private and split across owners. Federated learning trains a shared model without pooling raw data. Their framework, FedTPS, first learns to fake travel paths with a diffusion model.
Those fake paths then pad each client's local data. A new prediction model reads space and time links with attention tools. Tests use a large real ride sharing data set. FedTPS beat several federated rivals on global model quality. The design lets rivals in business still train together. Privacy stays intact while forecasts improve.
Catalog record
| CREATOR | Fermin Orozco, Hongkai Wen, Pedro Porto Buarque de Gusm̃ao, Johan Wahlström, Man Luo |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-08460