AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

Dynamic Personalized Federated Learning with Adaptive Differential Privacy

Xiyuan Yang, Wenke Huang, Mang Ye

NeurIPS 2023 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Bayesian personalized FL · 39 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Dynamic Fisher Personalization (ablation)DP-FedAvg base · FEMNIST eps=2average client accuracy (%): 71.92 -> 72.18Improved
add Dynamic Fisher Personalization (ablation)DP-FedAvg base · FEMNIST eps=4average client accuracy (%): 72.80 -> 74.56Improved
add Adaptive Constraint on top of DFP (ablation)FedDPA · FEMNIST eps=2average client accuracy (%): 72.18 -> 74.46Improved
add Adaptive Constraint on top of DFP (ablation)FedDPA · FEMNIST eps=16average client accuracy (%): 75.33 -> 76.99Improved
personalize by Fisher mask threshold tau=0.2 instead of tau=0 (all parameters personalized) (ablation)FedDPA · hyperparameter studyaccuracy (%): 39.72 -> 49.28Improved
vary tau between 0.1 and 1 (ablation)FedDPA · hyperparameter studyaccuracy (%): 47.97 - 49.28No clear effect
add L2 constraint on personalized updates lambda1=0.05 (ablation)FedDPA · hyperparameter studyaccuracy (%): 60.82 -> 62.64Improved
increase lambda1 from 0.05 to 0.1 or more (ablation)FedDPA · hyperparameter studyaccuracy (%): 62.64 -> 59.66Hurt
add bounded regularization on shared updates lambda2=0.1 (ablation)FedDPA · hyperparameter studyaccuracy (%): 58.34 -> 61.01Improved
increase lambda2 to 0.3 or 0.5 (ablation)FedDPA · hyperparameter studyaccuracy (%): 61.01 -> 45.66Hurt
FedDPA vs best baselineFedDPA · SVHN eps=6average client accuracy (%): +6.76Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy