The following projects are available for Master and Bachelor students. They are performed in close collaboration with an experienced member of the lab. Apply for a project by sending an email to the contact mentioned for the project.
You may also suggest new projects, ideally close enough to our ongoing, or previously completed projects. In that case, you will have to convince Anne-Marie Kermarrec that it is worthwhile, of reasonable scope, and that someone in the lab can mentor you!
Federated and Decentralized learning – model pruning
Contact: Drosis Odysseas <[email protected]>
Sparsification algorithms communicate a part of the model in decentralized and federated learning, but the training always happens on the full model. The aim of this project is to leverage the communication between different nodes to remove parameters from the model, hence reducing the size of the model. This would have an impact both on computation as well as communication.
Can we Really Spread Rumors Anonymously? (On the Differential Privacy of Gossip Protocols in Non-complete Graphs)
Contact: Kucherenko Anastasiia <[email protected]>
Gossip protocols (also called rumor spreading or epidemic protocols) are widely used to disseminate information in massive peer-to-peer networks. These protocols are often claimed to guarantee privacy because of the uncertainty they introduce on the node that started the dissemination. But is that claim really true? Can the source of a gossip safely hide in the crowd?
There are two sides of the project: 1) creating simulations of gossip propagation on a graph and getting empirical evaluations of privacy guarantees; 2) proving exact bounds on privacy using probability theory, graphs and algorithms tools.
Your awesome project here…