ORDERS: An Empirical Study of Norm-Rank Aggregation for Personalized Federated Learning

Name: ORDERS: An Empirical Study of Norm-Rank Aggregation for Personalized Federated Learning
Uploaded: 2026-10-17T23:27:27.441015+00:00
Duration: 338 s
Description: ORDERS-R1 tests a configuration combining shared models with client-specific predictors using norm-rank weighting for aggregation. It provides an auditable empirical characterization of this setup, showing performance on CIFAR-10 and Sent140, but does not establish universal guarantees regarding accuracy or robustness.

ORDERS-R1 tests a configuration combining shared models with client-specific predictors using norm-rank weighting for aggregation. It provides an auditable empirical characterization of this setup, showing performance on CIFAR-10 and Sent140, but does not establish universal guarantees regarding accuracy or robustness.

Open the full episode