E Pluribus Unum (Out of Many, One): Leveraging Federated Learning to Promote Deep Learning Model Development.

Abstract

The term federated is an adjective which traditionally refers to a country or organization set up as a single centralized unit within which each state or division keeps some internal autonomy [1]. Federated learning (FL) (also known as collaborative learning) is a machine learning technique that trains an algorithm across multiple decentralized facilities using local data, without directly exchanging that data. In essence, FL enables multiple, disparate groups to build a common machine learning model or algorithm without sharing data which ameliorates critical issues related to data privacy, data security, data access rights and access to heterogeneous data.