Skip navigation

Automated concept drift handling for fault prediction in edge clouds using reinforcement learning

Fault management systems that use real-time analytics based on Machine Learning help provide the reliability required in edge clouds. The highly dynamic traffic in edge clouds can cause concept drift – frequent changes in data distribution – which requires frequent adaptations of the ML model. We propose an automated concept drift handling framework for fault prediction in edge clouds. Reinforcement Learning is used to select the most appropriate drift adaptation method as well as the amount of data needed for adaptation, while considering the requirements of the edge cloud operator.
Research paper
#FaultPrediction #ReinforcementLearning #EricssonResearch

We implemented an edge cloud testbed and introduced infrastructure and network faults to it in the presence of abrupt and incremental concept drift. According to the obtained results, our proposed framework achieves up to 40% higher accuracy compared to a system without drift handling, and up to 13× and 30× less regret for selecting adaptation methods and amount of data, respectively, compared to other approaches.

Full abstract in IEEEXplore DOI: 10.1109/TNSM.2022.3153279 

     

Authors

Behshid Shayesteh, CIISE, Concordia University, Montréal, Canada 

Chunyan Fu, Ericsson Research

Amin Ebrahimzadeh, CIISE, Concordia University, Montréal, Canada 

Roch H. Glitho, CIISE, Concordia University, Montréal, Canada and University of Western Cape, Cape Town, South Africa 

Published in IEEE Transactions on Network and Service Management (Volume: 19, Issue: 2, June 2022)  

©2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse.