How to Evaluate Participant Contributions in Decentralized Federated Learning

TL;DR

This paper introduces TRIP-Shapley, a new method for evaluating participant contributions in decentralized federated learning, addressing challenges of model propagation and peer-to-peer exchanges without requiring model sharing.

Contribution

The paper presents TRIP-Shapley, a novel contribution evaluation method tailored for decentralized federated learning, overcoming limitations of existing centralized approaches.

Findings

TRIP-Shapley closely approximates ground-truth Shapley values.

It is scalable to large-scale scenarios.

It remains robust against dishonest clients.

Abstract

Federated learning (FL) enables multiple clients to collaboratively train machine learning models without sharing local data. In particular, decentralized FL (DFL), where clients exchange models without a central server, has gained attention for mitigating communication bottlenecks. Evaluating participant contributions is crucial in DFL to incentivize active participation and enhance transparency. However, existing contribution evaluation methods for FL assume centralized settings and cannot be applied directly to DFL due to two challenges: the inaccessibility of each client to non-neighboring clients' models, and the necessity to trace how contributions propagate in conjunction with peer-to-peer model exchanges over time. To address these challenges, we propose TRIP-Shapley, a novel contribution evaluation method for DFL. TRIP-Shapley formulates the clients' overall contributions by tracing the propagation of the round-wise local contributions. In this way, TRIP-Shapley accurately reflects the delayed and gradual influence propagation, as well as allowing a lightweight coordinator node to estimate the overall contributions without collecting models, but based solely on locally observable contributions reported by each client. Experiments demonstrate that TRIP-Shapley is sufficiently close to the ground-truth Shapley value, is scalable to large-scale scenarios, and remains robust in the presence of dishonest clients.