The First Workshop on Federated and Privacy Preserving AI for HPC
Session ChairsDirector of the National Center for Computational Sciences and the Oak Ridge Leadership Computing Facility
Event TypeWorkshop
Recorded
W
TimeMonday, 14 November 20228:30am - 12pm CST
LocationD220
DescriptionThe inundation of data generation technologies, along with progress in Artificial Intelligence (AI) and increasing privacy concerns, has prompted research into techniques for both federated and privacy preserving AI. Federated Learning (FL) allows multiple clients to collaborate in training AI models without sharing data while privacy preserving AI places further emphasis on protecting client data. To date, federated and privacy preserving AI are primarily driven by consumer demand for fast and accurate analysis on personal devices which may contain sensitive data. In the HPC domain, FL interest has grown in the areas of health analytics and coordination across experimental facilities. This latter form of FL poses new and unsolved problems. This workshop aims to highlight research in all aspects of federated and privacy preserving AI for HPC, machine learning, and scientific participants. Broad goals would be to consolidate the community around a core set of objectives and foster new collaborations.
Workshop Website
Workshop Website
Archive
view
Presentations
8:30am - 8:40am CST | Opening Remarks Presenter | |
8:40am - 9:20am CST | Privacy-Preserving Data Sharing and Analytics – An HPC Perspective Presenter | |
9:20am - 9:40am CST | Taking Federated Learning from Research to Real-World Deployment with NVIDIA FLARE Presenter | |
9:40am - 10:00am CST | Application of Privacy Preserving Federated Learning in Biomedical Applications – Lessons Learned from the PALISADE-X project Presenter | |
10:00am - 10:30am CST | Fed-PPAI - Morning break Presenter | |
10:30am - 10:50am CST | HPC at the FDA – Data Reuse Concepts for AI Presenter | |
10:50am - 11:10am CST | Federated Computations and the Digital Rights Movement Presenter | |
11:10am - 11:30am CST | Securing Federated Learning Capabilities within an Exascale Environment Presenter | |
11:30am - 12:00pm CST | Panel Discussion Moderator |