ALEXANDRIA, Va., Jan. 20 -- United States Patent no. 12,530,630, issued on Jan. 20, was assigned to Oracle International Corp. (Redwood City, Calif.).

"Hierarchical gradient averaging for enforcing subject level privacy" was invented by Virendra J. Marathe (Florence, Mass.) and Pallika Haridas Kanani (Westford, Mass.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Hierarchical gradient averaging is performed as part of training a machine learning model to enforce subject level privacy. A sample of data items from a training data set is identified and respective gradients for the data items are determined. The gradients are then clipped. Each subject's clipped gradients in the sample are averaged. A noise valu...