Data-Free Knowledge Distillation for Heterogeneous Federated Learning

Zhuangdi Zhu, Junyuan Hong, Jiayu Zhou
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:12878-12889, 2021.

Abstract

Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-zhu21b, title = {Data-Free Knowledge Distillation for Heterogeneous Federated Learning}, author = {Zhu, Zhuangdi and Hong, Junyuan and Zhou, Jiayu}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {12878--12889}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/zhu21b/zhu21b.pdf}, url = {https://proceedings.mlr.press/v139/zhu21b.html}, abstract = {Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.} }
Endnote
%0 Conference Paper %T Data-Free Knowledge Distillation for Heterogeneous Federated Learning %A Zhuangdi Zhu %A Junyuan Hong %A Jiayu Zhou %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-zhu21b %I PMLR %P 12878--12889 %U https://proceedings.mlr.press/v139/zhu21b.html %V 139 %X Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.
APA
Zhu, Z., Hong, J. & Zhou, J.. (2021). Data-Free Knowledge Distillation for Heterogeneous Federated Learning. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:12878-12889 Available from https://proceedings.mlr.press/v139/zhu21b.html.

Related Material