A method for joint training based on multi-party 3D printing database

A 3D printing and database technology, applied in electrical digital data processing, instruments, design optimization/simulation, etc., can solve problems such as application and development limitations, and achieve the effects of easy convergence, good practicability, and reduced number of rounds

Active Publication Date: 2022-07-15
CHENGDU AIRCRAFT INDUSTRY GROUP
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Industrial design, architecture, automobile, aerospace, dentistry, education fields, etc. are all applied, but its application and development are still limited by factors

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
  • A method for joint training based on multi-party 3D printing database
  • A method for joint training based on multi-party 3D printing database
  • A method for joint training based on multi-party 3D printing database

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] A method for joint training based on a multi-party 3D printing database, comprising a plurality of training members and a server, wherein the models of the training members are: Wi , the data for each training member is Xi , , labeled as y i ; the model of the server is W , and the server's model W Model with trained members Wi The network structure is consistent; including the following steps:

[0035] Step S100: in the first j During training iterations, train members i read Xi one of batch The data bi , and make the model Wi forward propagation of , to get the predicted label , and then according to the actual label y i , calculate the model Wi The loss function of , and then use the back-propagation algorithm to get the gradient matrix G i ;

[0036] Step S200: train members i pair gradient matrix G i The elements in are sorted from large to small according to the absolute value, and the first m elements are selected to obtain the correspo...

Embodiment 2

[0040] This embodiment is optimized on the basis of Embodiment 1. Before the iterative training, model initialization is performed: the server pairs the model W Initialize and send the initialization results to all training members, W i Initialize, determine the gradient upload scale coefficient α, attenuation coefficient ρ, learning rate γ.

[0041] Further, in the step S200, the statistical model is obtained W i The total number of elements in the M , calculate the number of gradient elements to be uploaded this time .

[0042] The other parts of this embodiment are the same as those of Embodiment 1, and thus are not repeated here.

Embodiment 3

[0044] This embodiment is optimized on the basis of Embodiment 1 or 2, and the correlation calculation in step S300 is as follows:

[0045]

[0046] in:

[0047]

[0048] in: D KL for KL Divergence,

[0049] P represents the data quality of each training member,

[0050] Q Indicates the data quality of all samples of the server.

[0051] The other parts of this embodiment are the same as the above-mentioned Embodiment 1 or 2, and thus are not repeated here.

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 discloses a method for joint training based on a multi-party 3D printing database. j During the training iteration process, the training members are obtained i The gradient matrix of G i ; train members i pair gradient matrix G i The elements in are sorted according to the absolute value, and the first m elements are selected to obtain the corresponding sparse matrix, and the filling element is 0; calculate the data quality of each training member Pi The correlation between the data quality Q corresponding to all samples of the server, and sort based on the correlation to obtain the training members participating in the current iteration; finally, use the sparse matrix of the training members participating in the current iteration to update the model of the server, and update the training accordingly member's model W i . The invention realizes joint training among multiple databases, and determines the training members of the current round of iterations based on the correlation, reduces the amount of data transmission in the joint training process, reduces the demand for data transmission bandwidth and investment cost, and has better performance. practicability.

Description

technical field [0001] The invention belongs to the technical field of joint processing of printing data, in particular to a method for joint training based on multi-party 3D printing databases. Background technique [0002] In the 1980s, 3D printing technology was born. 3D printing is not limited to traditional "removal" processing methods, and 3D printing is a bottom-up manufacturing method, also known as additive manufacturing technology, which realizes The establishment of mathematical models. 3D printing technology has received widespread attention since its birth, so it has developed rapidly. 3D printing technology has been in the spotlight in recent decades. Industrial design, architecture, automotive, aerospace, dentistry, education, etc. are all used, but their application and development are still limited by factors. [0003] During the implementation of 3D printing, because there are too many parameters related to 3D printing, it is impossible to exhaust all 3D...

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): G06F30/27G06N20/20G06F113/10
CPCG06F30/27G06N20/20G06F2113/10
Inventor 荣鹏高鹏高川云杜娟
Owner CHENGDU AIRCRAFT INDUSTRY GROUP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap