Byzantine-Robust Learning on Heterogeneous Datasets via Bucketing
In Byzantine robust distributed or federated learning, a central server wants to train a machine learning model over data distributed across multiple workers. However, a fraction of these workers may deviate from the prescribed algorithm and send arbitrary messages. While this problem has received s...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Artículo |
| Publicado: |
2020
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2006.09365 http://arxiv.org/abs/2006.09365 |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
Sea el primero en dejar un comentario!