Optimal client sampling for federated learning

W Chen, S Horvath, P Richtarik - arXiv preprint arXiv:2010.13723, 2020 - arxiv.org
It is well understood that client-master communication can be a primary bottleneck in
Federated Learning. In this work, we address this issue with a novel client subsampling …

[PDF][PDF] Optimal Client Sampling for Federated Learning

W Chen, S Horváth, P Richtárik - wenlin-chen.github.io
It is well understood that client-master communication can be a primary bottleneck in
federated learning (FL). In this work, we address this issue with a novel client subsampling …

Optimal Client Sampling for Federated Learning

W Chen, S Horváth, P Richtárik - Transactions on Machine Learning … - openreview.net
It is well understood that client-master communication can be a primary bottleneck in
federated learning (FL). In this work, we address this issue with a novel client subsampling …

Optimal Client Sampling for Federated Learning

W Chen, S Horvath, P Richtarik - arXiv e-prints, 2020 - ui.adsabs.harvard.edu
It is well understood that client-master communication can be a primary bottleneck in
Federated Learning. In this work, we address this issue with a novel client subsampling …

Optimal Client Sampling for Federated Learning

W Chen, S Horváth, P Richtárik - 2022 - repository.cam.ac.uk
It is well understood that client-master communication can be a primary bottleneck in
federated learning (FL). In this work, we address this issue with a novel client subsampling …

[PDF][PDF] Optimal Client Sampling for Federated Learning

W Chen, S Horvath, P Richtarik - 2020 - repository.kaust.edu.sa
It is well understood that client-master communication can be a primary bottleneck in
Federated Learning. In this work, we address this issue with a novel client subsampling …

Optimal Client Sampling for Federated Learning

W Chen, S Horváth, P Richtárik - openreview.net
It is well understood that client-master communication can be a primary bottleneck in
Federated Learning. In this work, we address this issue with a novel client subsampling …