Abstract: Since its inception by Google, Federated Learning (FL) has been instrumental in improving the performance of a wide range of applications. Android's Gboard for predictive text and Google Assistant are two of the most well-known and widely used FL-powered applications. FL is a configuration that enables on-device, collaborative Machine Learning. A diverse body of literature has investigated FL technical considerations, frameworks, and limitations, with several works presenting a survey of the prominent FL literature. Prior surveys, however, have focused on FL's technical considerations and challenges, and there has been a limitation in more recent work that presents a comprehensive overview of FL's status and future trends in applications and markets. We introduce the fundamentals of FL in this review, describing its underlying implementation of technologies, pros and cons, and recommendations along with privacy-preserving methods. More importantly, this work contributes to the understanding of a wide range of FL current applications and future trends in technology and markets today.
Keywords: Federated Learning, Cybersecurity, IoT, Decentralized networks, NIST.
|
DOI:
10.17148/IJARCCE.2023.12213
