Elastic Federated Learning with Kubernetes Vertical Pod Autoscaler for edge computing

Pham Khanh Quan, Taehong Kim

Abstract: Highlights•Federated Learning (FL): Decentralized machine learning adapting to edge computing.•Challenges in dynamic and resource-constraned in deploying FL applications.•Proposed Elastic FL framework with Kubernetes Vertical Pod Autoscaler (VPA).•VPA adjusts pod resources with FL workload using the historical and real-time data.•EFL framework enhances FL efficiency, addresses edge resource challenges.•Opening doors to more efficient scalable AI applications while ensuring data privacy.