Federated continual learning: A comprehensive survey on lifelong and privacy-preserving learning over distributed and non-stationary data

In three linesComprehensive survey on Federated Continual Learning (FCL), combining privacy-preserving federated learning with continual learning on non-stationary data. Identifies limitations of classical FL methods under evolving distributions and proposes a multi-dimensional taxonomy of FCL approaches for real-world applications (healthcare, industrial IoT, cybersecurity, smart cities).

Summary generated by Claude — human-verified