Federated Learning: Achieving Scalable and Privacy-Preserving Machine Learning

Federated Learning is a machine learning framework in which multiple clients engage in collaborative training of a model, while the training data remains distributed across the client devices. The training process is overseen by a centralised server. The utilisation of this approach effectively reduces numerous privacy risks and computation costs commonly associated with conventional, centralised machine learning techniques, wherein user data is transmitted to a central server for processing. This research paper is motivated by the increasing interest in the field of Federated Learning research, and aims to explore the advantages of this approach in comparison to conventional methods. The article additionally examines the obstacles and potential uses of Federated Learning, as this nascent framework garners increasing attention.