Capital one services, llc (20250139495). SELECTIVELY SHARED CONDENSED DATA FOR EFFICIENT FEDERATED LEARNING

SELECTIVELY SHARED CONDENSED DATA FOR EFFICIENT FEDERATED LEARNING

Organization Name

capital one services, llc

Inventor(s)

Michael Davis of Arlington VA US

Galen Rafferty of Mahomet IL US

Samuel Sharpe of Cambridge MA US

Brian Barr of Schenectady NY US

Jeremy Goodsitt of Champaign IL US

Taylor Turner of Richmond VA US

Kenny Bean of Herndon VA US

Owen Reinert of Queens NY US

Tyler Farnan of San Diego CA US

SELECTIVELY SHARED CONDENSED DATA FOR EFFICIENT FEDERATED LEARNING

This abstract first appeared for US patent application 20250139495 titled 'SELECTIVELY SHARED CONDENSED DATA FOR EFFICIENT FEDERATED LEARNING

Original Abstract Submitted

a method and related system for training a machine learning model using a federated learning structure by selectively sharing condensed data between devices includes operations to obtain datasets from client devices comprising a first client device and a second client device and updating a datasets subset to comprise a first dataset. the method further includes updating the datasets subset to comprise a second dataset based on a result indicating that a feature space distance between the first dataset and the second dataset satisfies a set of criteria and sending, to a third client device, the datasets subset comprising the first dataset and the second dataset. the method further includes obtaining, from the third client device, a set of model parameters that is derived from training based on the datasets subset and updating a server version of a machine learning model based on the set of model parameters.