Concept Drift Detection in Federated Networked Systems

TL;DR

This paper introduces a federated learning-based system for detecting concept drift in networked systems, using dimensionality reduction and clustering to identify affected nodes and maintain system performance.

Contribution

It presents a novel framework that leverages federated updates and clustering techniques to detect concept drift in distributed networks, especially in non-iid scenarios.

Findings

Effective detection of drifted nodes across various non-iid scenarios

Ability to identify drift at different stages and system exposure levels

Framework successfully isolates affected nodes in an intelligent transportation system

Abstract

As next-generation networks materialize, increasing levels of intelligence are required. Federated Learning has been identified as a key enabling technology of intelligent and distributed networks; however, it is prone to concept drift as with any machine learning application. Concept drift directly affects the model's performance and can result in severe consequences considering the critical and emergency services provided by modern networks. To mitigate the adverse effects of drift, this paper proposes a concept drift detection system leveraging the federated learning updates provided at each iteration of the federated training process. Using dimensionality reduction and clustering techniques, a framework that isolates the system's drifted nodes is presented through experiments using an Intelligent Transportation System as a use case. The presented work demonstrates that the proposed framework is able to detect drifted nodes in a variety of non-iid scenarios at different stages of drift and different levels of system exposure.