FLAME: Adaptive and Reactive Concept Drift Mitigation for Federated Learning Deployments

TL;DR

Compared to well-known lightweight mitigation methods, FLAME demonstrates superior performance in maintaining high F1 scores and reducing resource utilisation in large-scale IoT deployments, making it a promising approach for real-world applications.

Abstract

This paper presents Federated Learning with Adaptive Monitoring and Elimination (FLAME), a novel solution capable of detecting and mitigating concept drift in Federated Learning (FL) Internet of Things (IoT) environments. Concept drift poses significant challenges for FL models deployed in dynamic and real-world settings. FLAME leverages an FL architecture, considers a real-world FL pipeline, and proves capable of maintaining model performance and accuracy while addressing bandwidth and privacy constraints. Introducing various features and extensions on previous works, FLAME offers a robust solution to concept drift, significantly reducing computational load and communication overhead. Compared to well-known lightweight mitigation methods, FLAME demonstrates superior performance in maintaining high F1 scores and reducing resource utilisation in large-scale IoT deployments, making it a promising approach for real-world applications.

Figures (6)

• Figure 1: Overview of a typical FL lifecycle in FedOps for our scenario. It is distinguished in the "cloud" (handling the parameter server) and the "on-site" (handling the client and sensor nodes) deployments.
• Figure 2: A system diagram showing the operation of FLAME.
• Figure 3: a) Stability example for a model training on three different concepts in the MNIST-C dataset mnistC, b) an example of five concepts and the split of the newly created dataset.
• Figure 4: Small-factor experiment comparing static and dynamic thresholds.
• Figure 5: F1-Score of all different methods.