Unlock instant, AI-driven research and patent intelligence for your innovation.

Parallel training method and system for federated learning

A training method and federated technology, applied in machine learning, special data processing applications, instruments, etc., can solve problems such as graphics memory, computing power, and even quantity differences, and achieve the effect of solving hardware differences

Active Publication Date: 2021-04-02
CLUSTAR TECH LO LTD
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For example, the graphics cards owned by each entity will have large differences in video memory, computing power, and even the number

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
  • Parallel training method and system for federated learning
  • Parallel training method and system for federated learning
  • Parallel training method and system for federated learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] The purpose of this application is to provide a parallel training method and system for federated learning in order to solve the data correlation problem of vertical federated learning and the hardware difference problem between different entities. Wherein, the method includes: obtaining the intersection of the private data of the first participant and the private data of the second participant with the same flag; according to the mapping method determined based on the flag, the private data of the first participant The part corresponding to the intersection in the private data of the second participant and the part corresponding to the intersection in the private data of the second participant are both mapped to the associated data set; After grouping the hardware resources of the second participant, respectively obtain the hardware group of the first participant and the hardware group of the second participant; respectively obtain the associated data set and the hardwa...

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

This application relates to a parallel training method for federated learning. The method includes: obtaining the intersection of the private data of the first participant and the private data of the second participant with the same flag; obtaining a set of associated data according to a mapping method determined based on the flag; After grouping the hardware resources of the second party and the hardware resources of the second party, the hardware grouping of the first party and the hardware grouping of the second party are respectively obtained; A matching relationship and a second matching relationship between the associated data set and the hardware grouping of the second participant; and according to the first matching relationship and the second matching relationship, cooperatively mobilizing the hardware resources of the first participant and the hardware of the second participant resources for parallel training using linked datasets.

Description

technical field [0001] This application relates to the technical fields of data processing and privacy protection, and in particular to a parallel training method and system for federated learning. Background technique [0002] With the development of big data analysis and artificial intelligence technology, the demand for high-quality labeled data is increasing. For example, training a neural network and performing data mining require massive amounts of labeled data. These labeled data are often derived from the application data generated and accumulated in daily business activities. Application data is often scattered in the hands of different organizations and individuals. For example, transaction data is scattered in various financial institutions, and medical diagnosis data is scattered in various medical institutions. In addition, cross-industry and cross-field application data are often scattered. For example, social attribute data and e-commerce transaction data in...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06N20/00G06F16/901
CPCG06F16/9014G06N20/00
Inventor 彭瑞陆万航胡水海
Owner CLUSTAR TECH LO LTD