Communication-Efficient and Personalized Federated Lottery Ticket Learning

TL;DR

This paper introduces CELL, a federated lottery ticket learning algorithm that enhances communication efficiency and personalization by using broadcast transmission and user grouping, achieving higher accuracy with less communication.

Contribution

CELL is a novel federated lottery ticket learning method that reduces communication costs and mitigates stragglers through broadcast and user grouping strategies.

Findings

Achieves up to 3.6% higher personalized accuracy.

Reduces total communication cost by 4.3x.

Effective in CIFAR-10 classification tasks.

Abstract

The lottery ticket hypothesis (LTH) claims that a deep neural network (i.e., ground network) contains a number of subnetworks (i.e., winning tickets), each of which exhibiting identically accurate inference capability as that of the ground network. Federated learning (FL) has recently been applied in LotteryFL to discover such winning tickets in a distributed way, showing higher accuracy multi-task learning than Vanilla FL. Nonetheless, LotteryFL relies on unicast transmission on the downlink, and ignores mitigating stragglers, questioning scalability. Motivated by this, in this article we propose a personalized and communication-efficient federated lottery ticket learning algorithm, coined CELL, which exploits downlink broadcast for communication efficiency. Furthermore, it utilizes a novel user grouping method, thereby alternating between FL and lottery learning to mitigate stragglers. Numerical simulations validate that CELL achieves up to 3.6% higher personalized task classification accuracy with 4.3x smaller total communication cost until convergence under the CIFAR-10 dataset.