公开(公告)号：US11853891B2

公开(公告)日：2023-12-26

申请号：US16816153

申请日：2020-03-11

Applicant: SHARECARE AI, INC.

Inventor： Walter Adolf De Brouwer ,  Srivatsa Akshay Sharma ,  Neerajshyam Rangan Kashyap ,  Kartik Thakore ,  Philip Joseph Dow

IPC: G06K9/00 ,  G06N3/084 ,  G16H70/60 ,  G06N3/08 ,  G06V10/764 ,  G06V10/80 ,  G06V10/94 ,  G06V10/96

CPC classification number: G06N3/084 ,  G06N3/08 ,  G06V10/764 ,  G06V10/809 ,  G06V10/945 ,  G06V10/95 ,  G06V10/96 ,  G16H70/60

Abstract: Method and system with federated learning model for health care applications are disclosed. The system for federated learning comprises multiple edge devices of end users, one or more federated learner update repository, and one or more cloud. Each edge device comprises a federated learner model, configured to send tensors to federated learner update repository. Cloud comprises a federated learner model, configured to send tensors to federated learner update repository. Federated learner update repository comprises a back-end configuration, configured to send model updates to edge devices and cloud.

公开(公告)号：US11811794B2

公开(公告)日：2023-11-07

申请号：US17319025

申请日：2021-05-12

Applicant: Sharecare AI, Inc.

Inventor： Gabriel Gabra Zaccak ,  William Hartman ,  Andrés Rodriguez Esmeral ,  Devin Daniel Reich ,  Marina Titova ,  Brett Robert Redinger ,  Philip Joseph Dow ,  Satish Srinivasan Bhat ,  Walter Adolf De Brouwer ,  Scott Michael Kirk

IPC: H04L9/40 ,  G06N20/20

CPC classification number: H04L63/1416 ,  G06N20/20

Abstract: The technology disclosed provides systems and methods related to preventing exfiltration of training data by feature reconstruction attacks on model instances trained on the training data during a training job. The system comprises a privacy interface that presents a plurality of modulators for a plurality of training parameters. The modulators are configured to respond to selection commands via the privacy interface to trigger procedural calls. The procedural calls modify corresponding training parameters in the plurality of training parameters for respective training cycles in the training job. The system comprises a trainer configured to execute the training cycles in dependence on the modified training parameters. The trainer can determine a performance accuracy of the model instances for each of the executed training cycles. The system comprises a differential privacy estimator configured to estimate a privacy guarantee for each of the executed training cycles in dependence on the modified training parameters.