… a personalizedfederatedlearning setting involving local and global models subject to user-level (joint) differential privacy. While learning … on personalizedfederatedlearning algorithms …
Personalizedfederatedlearning (PFL), as a novel federatedlearning (FL) paradigm, is capable of generating personalized models for heterogenous clients. Combined with a meta-…
X Yang, W Huang, M Ye - Advances in Neural Information …, 2023 - proceedings.neurips.cc
… We introduce an innovative approach to personalizedfederatedlearning. Our method, based on layer-wise Fisher information values, allows for dynamic personalization that surpasses …
… personalization strategies may be preferred. In this paper, we consider the federated meta-learning problem of learningpersonalization … the batch-norm and learning rate parameters …
… We study federatedlearning from a multi-task learning perspective by leveraging both shared representation and inter-client classifier collaboration. Specifically, we make use of the …
B Sun, H Huo, Y Yang, B Bai - Advances in Neural …, 2021 - proceedings.neurips.cc
… Based on such generous observations, we propose a PersonalizedFederatedLearning (PFL) method from the idea of client-agnostic and client-specific initialization. Initialization is …
J Wang, G Xu, W Lei, L Gong… - IEEE Transactions on …, 2022 - ieeexplore.ieee.org
… As a possible solution, the existing personalizedfederatedlearning relies too much on a … whole federation effect. In this article, we design a cognitive personalizedfederatedlearning (…
H Li, Z Cai, J Wang, J Tang, W Ding… - … and learning systems, 2023 - ieeexplore.ieee.org
… of the transformer model in federatedlearning settings. To … based federatedlearning framework that learns personalized … personalization mechanism that maintains personalized self-…
… A widely recognized difficulty in federatedlearning arises … unrelated distributions, and personalization is, therefore, … excess risks of personalizedfederatedlearning with a smooth, …