| CPC H04L 63/1416 (2013.01) [G06N 20/00 (2019.01); H04L 63/20 (2013.01)] | 18 Claims |

|
1. A method, comprising:
performing a federated learning process comprising:
receiving, at a central node, a respective gradient from each node of a cluster, and each of the gradients comprises a respective update to a global machine learning model maintained at the central node;
aggregating, by the central node, the gradients to obtain an aggregated gradient for the cluster, and the aggregated gradient is part of a list of aggregated gradients;
running, by the central node, a robust aggregation operation on the aggregated gradients in the list to obtain an outlier score for the cluster;
when the outlier score equals or exceeds a specified value, or falls within a specified range of values, identifying the cluster as an outlier, and identifying nodes within the cluster as suspicious;
identifying one of the suspicious nodes as a Byzantine node; and
determining that the Byzantine node is carrying out a Byzantine attack on the central model, wherein the federated learning process is allowed to continue [1] notwithstanding the determination that the Byzantine attack is being carried out, and [2] without modifying an ongoing re-clustering operation that does not involve the Byzantine node.
|