A Beginner's Guide to Federated Learning

The increasing prevalence of Internet of Things devices produces substantial volumes of data, thereby demanding the development of advanced machine learning (ML) models to maintain the data securely. Federated learning (FL) is such an advanced distributed ML paradigm that enables multiple devices or data centers to collaboratively train an ML model without the need to share the raw data. In recent years, it has got considerable interest because of its capability to overcome the challenges of traditional centralized ML models, such as data privacy and scalability. Even though there is such growing interest in FL, there are merely very few resources available in the literature to fundamentally understand and implement the FL concepts. In light of this, the objective of this paper is to provide basic guidance that is essential for beginners to develop FL-based solutions. This paper provides a clear intuition behind the FL, which includes a discussion on different types (horizontal FL, vertical FL, and federated transfer learning) and federated averaging (FedAvg) algorithm. It also highlights various key applications of FL. Finally, the paper emphasizes a few research directions that facilitate the development of FL.