Abstract
News editors obtained the following quote from the background information suppli ed by the inventors:“Federated learning is a machine learning technique that di ffers from traditional centralized machinelearning technique, where all relevan t data is centralized on a single computing device for training toproduce a mod el, the federated learning may train the machine learning models across multiple deviceswithout the need to transfer training data. The federated learning adop ts multi-node computation ofthe distributed system that distributes the work of model-training to multiple nodes. Distributed systemsprimarily deal with a lar ge amount of data and allocate the computing power to relevant units. However, the federated learning emphasizes on the protection of the data privacy, so the t raining data does not needto be centralized on a single server for distribution and computation. Instead, each device may individuallybe used for training the model.”
NETL
NSTL