Bandit-based Communication-Efficient Client Selection Strategies for Federated Learning

Due to communication constraints and intermittent client availability in federated learning, only a subset of clients can participate in each training round. While most prior works assume uniform and unbiased client selection, recent work on biased client selection has shown that selecting clients with higher local losses can improve error convergence speed. However, previously proposed biased selection strategies either require additional communication cost for evaluating the exact local loss or utilize stale local loss, which can even make the model diverge. In this paper, we present a bandit-based communication-efficient client selection strategy UCB-CS that achieves faster convergence with lower communication overhead. We also demonstrate how client selection can be used to improve fairness.

[1]  Anit Kumar Sahu,et al.  Federated Optimization in Heterogeneous Networks , 2018, MLSys.

[2]  Ashutosh Sabharwal,et al.  An Axiomatic Theory of Fairness in Network Resource Allocation , 2009, 2010 Proceedings IEEE INFOCOM.

[3]  Roland Vollgraf,et al.  Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms , 2017, ArXiv.

[4]  Sebastian U. Stich,et al.  Local SGD Converges Fast and Communicates Little , 2018, ICLR.

[5]  Tzu-Ming Harry Hsu,et al.  Measuring the Effects of Non-Identical Data Distribution for Federated Visual Classification , 2019, ArXiv.

[6]  Mung Chiang,et al.  Multiresource allocation: fairness-efficiency tradeoffs in a unifying framework , 2013, TNET.

[7]  Jianyu Wang,et al.  Cooperative SGD: A unified Framework for the Design and Analysis of Communication-Efficient SGD Algorithms , 2018, ArXiv.

[8]  Richard Nock,et al.  Advances and Open Problems in Federated Learning , 2019, Found. Trends Mach. Learn..

[9]  Samarth Gupta,et al.  Correlated Multi-Armed Bandits with A Latent Random Source , 2018, ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[10]  Mehryar Mohri,et al.  Agnostic Federated Learning , 2019, ICML.

[11]  Anuj Kumar,et al.  Active Federated Learning , 2019, ArXiv.

[12]  Sébastien Bubeck,et al.  Regret Analysis of Stochastic and Nonstochastic Multi-armed Bandit Problems , 2012, Found. Trends Mach. Learn..

[13]  Raj Jain,et al.  A Quantitative Measure Of Fairness And Discrimination For Resource Allocation In Shared Computer Systems , 1998, ArXiv.

[14]  Tian Li,et al.  Fair Resource Allocation in Federated Learning , 2019, ICLR.

[15]  Shenghuo Zhu,et al.  Parallel Restarted SGD for Non-Convex Optimization with Faster Convergence and Less Communication , 2018, ArXiv.

[16]  Aurélien Garivier,et al.  On Upper-Confidence Bound Policies for Non-Stationary Bandit Problems , 2008, 0805.3415.

[17]  Shenghuo Zhu,et al.  Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning , 2018, AAAI.

[18]  Jianyu Wang,et al.  Client Selection in Federated Learning: Convergence Analysis and Power-of-Choice Selection Strategies , 2020, ArXiv.

[19]  Blaise Agüera y Arcas,et al.  Communication-Efficient Learning of Deep Networks from Decentralized Data , 2016, AISTATS.

[20]  Samarth Gupta,et al.  Multi-Armed Bandits with Correlated Arms , 2019, ArXiv.