Data-Free Evaluation of User Contributions in Federated Learning

TL;DR

This paper introduces PCA, a novel test dataset-free method for evaluating user contributions in federated learning, enabling better incentives and detection of malicious users, with demonstrated effectiveness on real datasets.

Contribution

The paper proposes PCA, a peer prediction-based approach for user contribution evaluation in federated learning without requiring a test dataset, and integrates it into Fed-PCA with incentive guarantees.

Findings

Fed-PCA outperforms FedAvg in accuracy

PCA effectively incentivizes truthful user behavior

Method validated on MNIST and industrial datasets

Abstract

Federated learning (FL) trains a machine learning model on mobile devices in a distributed manner using each device's private data and computing resources. A critical issues is to evaluate individual users' contributions so that (1) users' effort in model training can be compensated with proper incentives and (2) malicious and low-quality users can be detected and removed. The state-of-the-art solutions require a representative test dataset for the evaluation purpose, but such a dataset is often unavailable and hard to synthesize. In this paper, we propose a method called Pairwise Correlated Agreement (PCA) based on the idea of peer prediction to evaluate user contribution in FL without a test dataset. PCA achieves this using the statistical correlation of the model parameters uploaded by users. We then apply PCA to designing (1) a new federated learning algorithm called Fed-PCA, and (2) a new incentive mechanism that guarantees truthfulness. We evaluate the performance of PCA and Fed-PCA using the MNIST dataset and a large industrial product recommendation dataset. The results demonstrate that our Fed-PCA outperforms the canonical FedAvg algorithm and other baseline methods in accuracy, and at the same time, PCA effectively incentivizes users to behave truthfully.