logo

For the Inventor, By the Inventor

The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.

The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.

Patent No.:

US 12762472 B1

PDFFull Text
Date of Patent:
Oct. 13, 2026

Filed:

Nov. 17, 2022

Model training method and apparatus for federated learning, device, and storage medium

Applicant:

Tencent Technology (Shenzhen) Company Limited, Shenzhen, CN;

Inventors:

Yong Cheng, Shenzhen, CN;

Yangyu Tao, Shenzhen, CN;

Shu Liu, Shenzhen, CN;

Jie Jiang, Shenzhen, CN;

Yuhong Liu, Shenzhen, CN;

Peng Chen, Shenzhen, CN;

Assignee:

TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED, Shenzhen, CN;

Attorney:

Crowell & Moring LLP

Primary Examiner:

Abdullah Al Kawsar

Int. Cl.
CPC ...
G06N 20/00 (2019.01); G06F 7/32 (2006.01); G06F 7/499 (2006.01); G06F 9/54 (2006.01); G06N 3/045 (2023.01); G06N 3/084 (2023.01);
U.S. Cl.
CPC ...
G06N 20/00 (2019.01); G06F 7/32 (2013.01); G06F 7/49947 (2013.01); G06F 9/54 (2013.01); G06N 3/045 (2023.01); G06N 3/084 (2013.01);
Abstract

A model training method and apparatus for federated learning, a device and a storage medium are provided, which belong to the technical field of machining learning. The method includes: generating an iscalar operator based on a (t−1)round of training data and a tround of training data (); transmitting an ifusion operator to a next node device based on the iscalar operator (); determining an isecond-order gradient descent direction of an isub-model based on an acquired second-order gradient scalar, an imodel parameter and an ifirst-order gradient; and updating the isub-model based on the isecond-order gradient descent direction to obtain a model parameter of the isub-model during a (t+1)round of iterative training.

Find Patent Forward Citations

Cheng, Yong. (2026). Model training method and apparatus for federated learning, device, and storage medium (U.S. Patent No. 12762472). U.S. Patent and Trademark Office. https://idiyas.com/patent/badge/12762472

Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile

Loading…