Discover published sets by community

The safeguarding of sensitive information in federated learning by employing techniques such as encryption, differential privacy, and secure multiparty computation to protect against data breaches and privacy leaks.

The method of combining the updates from multiple local models to improve a shared global model without compromising the data's privacy.

The process of training a model on a user's device using their own data without sharing the data with a central server or other devices.

The optimization of the data transmission process between local devices and the central server to reduce communication load and latency in federated learning.

An algorithm used in federated learning where the server computes the average of the locally computed updates on the client model weights, which is then used to update the global model.

Adapting a federated learning model to cater to the individual preferences or behaviors of users based on their unique data.

The policy used to determine which devices or clients are chosen to participate in a round of federated learning training to optimize resource usage and learning efficiency.