A sandbox website to see how a federated learning ML model is trained accross multiple servers without leaking data
Imagine 3 different hospitals, all wanting to create an AI model together. It is not realistic or secure to share data to create a model. Federated learning enables multiple servers to collaborate on a model, without sending data away from local.
iPhone data can be pooled together to create AI about health, SIRI, etc, without sending your personal data away from your phone. Banks can create better fraud detection by collaborating. Anything in which data privacy is key can be benefited by federated learning.
YouTube Video
