tinyML Talks: Unsupervised Federated Learning

“Unsupervised Federated Learning”

Ranjitha Prasad
Assistant Professor
IIIT Delhi

Modern day applications such as autonomous vehicles, IoT, smart grids, etc., generate massive amounts of data at the edge. Federated Learning (FL) enables machine learning without having to transfer data from edge devices to any untrusted third party. A fundamental challenge in federated supervised learning is ensuring that data at the edge is annotated. This talk gives a general overview of federated learning with specific focuss on federated learning techniques that utilise the unannotated data at the edge for learning a global model.

source