"Besides, intuitively, keeping the data at its source is valuable in any privacy-preserving applications, especially when applied in healthcare or on confidential data in business and government."

"Initially proposed in 2015, federated learning is an algorithmic solution that enables the training of ML models by sending copies of a model to the place where data resides  and performing training at the edge, thereby eliminating the necessity to move large amounts of data to a central server for training purposes."

"The data remains at its source devices, a.k.a. the clients, which receive a copy of the global model from the central server. This copy of the global model is trained locally with  the data of each device. The model weights are updated via local training, and then the local copy is  sent back to the central server."

"Once the server receives the updated model, it proceeds to aggregate the updates, improving the global model without revealing any of the private data on which it was trained."

"One of the first applications of FL was to improve word recommendation in Google's Android keyboard without uploading the data, i.e. a user’s text, to the cloud."

"More recently, Apple has detailed how it employs federated learning to improve Siri's voice recognition."

"The cost for implementing federated learning is higher than collecting the information and processing it centrally, especially during the early phases of R&D when the training method and process are still being iterated on."

"Researchers can train models using private and sensitive data without having to worry about handling the data - the data remains on the device and only learned model updates are transferred between the lab and the data owners."

"Requires data owners to perform computations on the device that holds data - for some devices with limited computation capacity this may not be possible or economic."

"Implementing FL is not enough to guarantee privacy, as model updates may include traces that can be used to infer private and sensitive information, thus requiring mixing with other techniques."

"To get started we will use the classical MNIST data set that will stand in for our clients' data,"

"As we mentioned above the larger vision for the technology goes beyond any single application or service."

"Data owners and data scientists can connect on the platform, where the data owners can feel safe in the knowledge that their data will never leave their node, and data scientists can perform their analysis without infringing on anyone's privacy rights."