Adaptive Low-Rank Optimisation for Federated Learning

This thesis explored how low-rank capacity can adapt to heterogeneous clients and changing optimisation needs instead of remaining fixed throughout federated training.