US Patent Issued to Google on Oct. 6 for "Learning with label differential privacy via projections" (American, Thai Inventors)

ALEXANDRIA, Va., Oct. 6 -- United States Patent no. 12,755,205, issued on Oct. 6, was assigned to Google LLC (Mountain View, Calif.).

"Learning with label differential privacy via projections" was invented by Badih Ghazi (Cupertino, Calif.), Yangsibo Huang (Princeton, N.J.), Chiyuan Zhang (Mountain View, Calif.), Pasin Manurangsi (Bangkok), Pritish Kamath (Mountain View, Calif.) and Shanmugasundaram Ravikumar (Piedmont, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Aspects of the disclosure are directed to implementing a projection-based stochastic gradient descent technique that maintains label differential privacy when training one or more machine learning models. The technique includes denoising g...