FedGP: Correlation-Based Active Client Selection for Heterogeneous Federated Learning

03/24/2021
by   Minxue Tang, et al.
0

Client-wise heterogeneity is one of the major issues that hinder effective training in federated learning (FL). Since the data distribution on each client may differ dramatically, the client selection strategy can largely influence the convergence rate of the FL process. Several recent studies adopt active client selection strategies. However, they neglect the loss correlations between the clients and achieve marginal improvement compared to the uniform selection strategy. In this work, we propose FedGP – a federated learning framework built on a correlation-based client selection strategy, to boost the convergence rate of FL. Specifically, we first model the loss correlations between the clients with a Gaussian Process (GP). To make the GP training feasible in the communication-bounded FL process, we develop a GP training method utilizing the historical samples efficiently to reduce the communication cost. Finally, based on the correlations we learned, we derive the client selection with an enlarged reduction of expected global loss in each round. Our experimental results show that compared to the latest active client selection strategy, FedGP can improve the convergence rates by 1.3∼2.3× and 1.2∼1.4× on FMNIST and CIFAR-10, respectively.

READ FULL TEXT

page 6

page 12

research
03/30/2023

DPP-based Client Selection for Federated Learning with Non-IID Data

This paper proposes a client selection (CS) method to tackle the communi...
research
12/19/2022

Adaptive Control of Client Selection and Gradient Compression for Efficient Federated Learning

Federated learning (FL) allows multiple clients cooperatively train mode...
research
08/10/2022

Fast Heterogeneous Federated Learning with Hybrid Client Selection

Client selection schemes are widely adopted to handle the communication-...
research
01/15/2022

Variance-Reduced Heterogeneous Federated Learning via Stratified Client Selection

Client selection strategies are widely adopted to handle the communicati...
research
11/19/2021

Client Selection in Federated Learning based on Gradients Importance

Federated learning (FL) enables multiple devices to collaboratively lear...
research
11/17/2020

Stochastic Client Selection for Federated Learning with Volatile Clients

Federated Learning (FL), arising as a novel secure learning paradigm, ha...
research
06/20/2021

Is Shapley Value fair? Improving Client Selection for Mavericks in Federated Learning

Shapley Value is commonly adopted to measure and incentivize client part...

Please sign up or login with your details

Forgot password? Click here to reset