Efficient Client Selection in Federated Learning

TL;DR

This paper introduces an adaptive client selection framework for federated learning that enhances privacy, robustness, and efficiency, leading to significant accuracy and training time improvements in network anomaly detection tasks.

Contribution

It presents a novel client selection method combining differential privacy and fault tolerance, optimizing performance and system constraints in federated learning.

Findings

Improves accuracy by 7% in anomaly detection.

Reduces training time by 25%.

Enhances robustness with minimal performance loss.

Abstract

Federated Learning (FL) enables decentralized machine learning while preserving data privacy. This paper proposes a novel client selection framework that integrates differential privacy and fault tolerance. The adaptive client selection adjusts the number of clients based on performance and system constraints, with noise added to protect privacy. Evaluated on the UNSW-NB15 and ROAD datasets for network anomaly detection, the method improves accuracy by 7% and reduces training time by 25% compared to baselines. Fault tolerance enhances robustness with minimal performance trade-offs.