Comments on "Federated Learning with Differential Privacy: Algorithms and Performance Analysis"

TL;DR

This paper corrects the convergence upper bound for the NbAFL algorithm in federated learning with differential privacy, addressing inaccuracies in prior theoretical analysis.

Contribution

It provides the accurate convergence upper bound for NbAFL, clarifying the theoretical understanding of its performance in federated learning.

Findings

Corrected the convergence upper bound for NbAFL

Clarified theoretical performance limits of the algorithm

Enhanced understanding of differential privacy in federated learning

Abstract

In the paper by Wei et al. ("Federated Learning with Differential Privacy: Algorithms and Performance Analysis"), the convergence performance of the proposed differential privacy algorithm in federated learning (FL), known as Noising before Model Aggregation FL (NbAFL), was studied. However, the presented convergence upper bound of NbAFL (Theorem 2) is incorrect. This comment aims to present the correct form of the convergence upper bound for NbAFL.