Presenter: Yuandou Wang, Multiscale Networked (MNS) Systems, University of Amsterdam, the Netherlands
Co-authors: Chronis Kontomaris, Zhiming Zhao
Schedule:
Session 1 (Online only) Wednesday 8 May, Americas friendly time,
and
Session 4 (Amsterdam / Online hybrid) Wednesday 15 May, EMEA Friendly time
In this presentation, we start from the background of federated learning operations (FLOps) concerning the diversity and customization in the federated learning research and development process and identify the research gap. Then we center around the main research question: How do we achieve a highly automated FL process while supporting diverse FL scenarios? We proposed our solution to this question. We introduce a novel CWL-based method, named CWL_FLOps, that leverages common workflow language to automate the federated learning operations at scale and present the evaluation process.
Please leave your questions for the presenter below!
As an alternative to YouTube, this presentations is also available on ConfTube