Coding Analysis and Integrated Learning Experiments Augmented with Different Gossipy and Privacy Processes
In this tutorial, we examine how clustered learning behaves when the central clustering server is removed and replaced with a peer-to-peer distributed gossip mechanism. We implement both centralized FedAvg and Decentralized Gossip Federated Learning from scratch and introduce client-side partition privacy by injecting limited noise into local model updates. Using controlled experiments on non-IID MNIST … Read more