Would you be interested on a talk on decentralized AI?
- Interested
- Not Interested
0
voters
Professor Nic Lane will be visiting NZ between April 5th to 14th and has offered to do a talk for Health NZ on deploying a Federated Learning Framework.
-
Federated Learning is a machine learning approach where a model is trained across multiple decentralized devices or servers with local data, without exchanging these data samples.
- Involves sending the model to the data source, training it locally, and only sharing model updates
- Particularly beneficial for us in healthcare where data sensitivity, sovereignty and privacy are paramount.
-
Nic Lane is a Professor at Cambridge and co-founder of Kiwi AI start-up Flower Labs – see: NZH article
-
Flower Labs has developed the Flower federated learning framework enabling collaborative, open and distributed AI used by NHS, Siemens Health, Harvard, MIT, amongst others.
- Website: www.flower.ai
- FSI ’24 talk: Introducing FlowerLLM (Flower AI Summit 2024)
- Milestones:
- FL can train true foundation models
- Only group so far to show this can be done with sizes >1B parameters (Google DeepMind = 0.4B) and is currently training even larger ones as proof points for early adopting organizations.
-
Benefits for Health NZ of federated learning include:
- Enabling collaborative, privacy-preserving machine learning across diverse datasets.
- Potentially improving diagnostics and treatment while maintaining data security and regulatory compliance.
- Facilitating cost-effective, efficient model development and innovation in public health research without compromising patient privacy.
CC: @chris.paton