Check patentability & draft patents in minutes with Patsnap Eureka AI!

Electronic equipment and local model aggregation method and device of joint learning engine

A local model and learning engine technology, applied in the field of artificial intelligence, can solve problems such as disconnection on the user side and failure to perform aggregation functions on the server side, and achieve the effect of improving aggregation efficiency

Pending Publication Date: 2022-07-19
ENNEW DIGITAL TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In view of this, an embodiment of the present invention provides a local model aggregation method and device for an electronic device and a joint learning engine to solve the problem that the service side cannot normally perform the aggregation function due to an abnormality such as a disconnection on the user side during the joint learning process

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Electronic equipment and local model aggregation method and device of joint learning engine
  • Electronic equipment and local model aggregation method and device of joint learning engine
  • Electronic equipment and local model aggregation method and device of joint learning engine

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] figure 2 This is a flowchart of the local model aggregation method of the joint learning engine provided in the first embodiment of the present invention.

[0028] Among them, this embodiment describes the local model aggregation method of the joint learning engine from the service side, combined with figure 2 As shown, the method specifically includes steps S101-S104.

[0029] Step S101: Initialize the joint learning task, where the initializing the joint learning task includes: initializing the number of aggregations and initializing the global model.

[0030] The joint learning task is a joint learning task established on the data of different clients according to the joint learning algorithm, wherein the joint learning task includes the setting and initialization of the number of iterations (or rounds) of joint learning, and the number of iterations That is, the number of aggregations. The number of aggregations may be the number of times that all the clients p...

Embodiment 2

[0047] Figure 4 This is a structural diagram of the local model aggregation apparatus of the joint learning engine provided in the second embodiment of the present invention. Among them, this embodiment is based on the same inventive concept as the first embodiment, aiming at figure 2 The steps in the shown method embodiments correspond to specific device embodiments one-to-one.

[0048] like Figure 4 As shown, the local model aggregation device 400 of the joint learning engine specifically includes: a task initialization module 401, which is used to initialize the joint learning task, and the initial joint learning task includes: initializing the number of aggregations and initializing the global model; the local model receiving module 402, for checking whether a client uploads a new partial model according to the joint task; the partial model aggregation module 403, for when receiving a new partial model, by combining the new partial model with the previous aggregation ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention is suitable for the technical field of artificial intelligence, and provides an electronic device and a local model aggregation method and device for a joint learning engine, and the method comprises the steps: initializing a joint learning task, which comprises the initialization of the number of aggregation times and the initialization of a global model; according to the joint task, checking whether a client uploads a new local model or not; when a new local model is received, the global model is aggregated and updated by performing weighted summation on the new local model and the global model aggregated last time, the weighting coefficient of the new local model comprises the reciprocal of the number of times of aggregation this time, and the weighting coefficient of the new local model comprises the reciprocal of the number of times of aggregation this time; the weighting coefficient of the global model after the previous polymerization comprises the ratio of the previous polymerization times to the current polymerization times. According to the method, only a new local model needs to be uploaded, and the joint learning engine carries out aggregation, so that joint learning is not influenced under some abnormal conditions, and the aggregation efficiency can be effectively improved.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, and in particular relates to a local model aggregation method and device of an electronic device and a joint learning engine. Background technique [0002] During the training of machine learning algorithms using federated learning, if a user node is disconnected on the user side, the federated learning will be abnormal and the server side cannot complete the aggregation task. In view of this situation found in the prior art, how to overcome the problem that the service side cannot normally perform the aggregation function due to an abnormality such as a disconnection on the user side during the joint learning process is a technical problem that needs to be solved at present. SUMMARY OF THE INVENTION [0003] In view of this, the embodiments of the present invention provide a local model aggregation method and apparatus for an electronic device and a joint learning engine, so as...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06N20/00G06N3/08
CPCG06N20/00G06N3/082
Inventor 张敏高庆
Owner ENNEW DIGITAL TECH CO LTD
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com