Federated Learning With Health Data: Privacy-Preserving AI
Impossibile aggiungere al carrello
Rimozione dalla Lista desideri non riuscita.
Non è stato possibile aggiungere il titolo alla Libreria
Non è stato possibile seguire il Podcast
Esecuzione del comando Non seguire più non riuscita
Federated Learning With Health Data: Privacy-Preserving AI
-
Letto da:
-
Di:
In this edition of the LEAPcast we talk to Dr Huw Day, Huw is a post-doctoral research associate in digital health at the VIVO Hub for Enhanced Independent Living. Having completed his PhD in Mathematics here at the University of Bristol, Huw began postdoctoral work in data science and machine learning. He also has a keen interest in data ethics, and advocate for outreach. You can find Huw via his LinkedIn profile here or email him on huw.day@bristol.ac.uk.
We discuss federated learning and why it matters for healthcare and other privacy-sensitive domains. Federated learning addresses situations where data can’t be centralised. Instead models are sent out to clients (e.g., hospitals or homes), trained locally, and then aggregated on a central server. Huw discusses real deployments, key barriers such as governance, and privacy limits. He also describes Vivo’s work using federated learning to train models that control soft wearable exosuits from sensor data, while keeping sensitive in-home data local.
Other useful links mentioned in this episode:
https://www.thevivohub.com/
https://docs.isambard.ac.uk/access/
https://www.linkedin.com/in/huw-day/