Federated Learning in Distributed Cloud Systems: Enhancing Privacy and Scalability for Machine Learning in Edge Computing

Abstract

In the era of edge computing, where data is generated and processed at the network's edge, ensuring privacy and scalability in machine learning models is paramount. Federated Learning (FL) addresses these challenges by allowing multiple edge devices to collaboratively train models without sharing raw data. This paper investigates the implementation of FL in distributed cloud systems, highlighting its role in preserving data privacy and improving scalability. We analyze various FL algorithms, such as Federated Averaging (FedAvg) and Hybrid Federated Dual Coordinate Ascent (HyFDCA), assessing their effectiveness in edge computing contexts. Additionally, we explore techniques like inverse distance aggregation to handle non-IID data distributions and discuss the trade-offs between communication and computation in FL frameworks. Through comprehensive analysis and experimentation, this study provides insights into optimizing FL for edge computing, paving the way for more secure and scalable machine learning applications in distributed cloud environments.