Simulation deep learning construction method based on SDL model

A technology of deep learning and construction methods, applied in the direction of neural learning methods, biological neural network models, neural architectures, etc., can solve problems such as non-existence, achieve high performance, low import cost, and easy popularization

Pending Publication Date: 2022-05-27
顾泽苍
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But it does not have the feature of deep learning function mapping model that can expand the interval between feature vectors

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
  • Simulation deep learning construction method based on SDL model
  • Simulation deep learning construction method based on SDL model
  • Simulation deep learning construction method based on SDL model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0083] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings, but the embodiments of the present invention are illustrative rather than restrictive.

[0084] First, some new definitions, new concepts and new formulas required by the present invention are introduced

[0085] Self-discipline learning SDL (Self-Discepline Learning) model

[0086] Let the probability space:

[0087] 【Formula 1】

[0088]

[0089] There is an initial area, the center value of this area And the variance of the Gaussian distribution calculated with this central value as the scale of the initial maximum probability by is the central value, on a probability scale The following iterations are performed for the benchmark:

[0090] [Formula 2]

[0091]

[0092]

[0093]

[0094] As a result of the n iterations, a maximum probability value close to the parent in the above probability space must be obtained. max...

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 relates to a simulation deep learning construction method based on an SDL model in the field of information processing, which is characterized in that feature values of all training data of an identified object are mapped to a data set layer through a mapping function, or a Gaussian distribution result is mapped to the data set layer through the mapping function. The method is characterized in that for data training of an identified object, a training effect of a deep learning function mapping model can be achieved only through an algorithm of a mapping function without a combination method, a black box problem is avoided, large hardware support is not needed, the function mapping model can realize high-precision identification capability of deep learning, and the identification efficiency is improved. And the generalization ability beyond deep learning can be obtained by adding the Gaussian distribution model, and the annotation of big data is not needed, so that the performance is high, the import cost is low, and large-scale popularization is facilitated.

Description

【Technical field】 [0001] The invention belongs to a construction method of simulated deep learning based on an SDL model in the field of artificial intelligence. 【Background technique】 [0002] The "deep learning" proposed by Hinton's team at the University of Toronto in Canada (Non-Patent Document 1) has achieved excellent results in the image classification test data set of IMAGENET, which has attracted the attention of the world, thus setting off the climax of this artificial intelligence. Many researchers are trying to use "deep learning" models to control self-driving cars. A representative method is "learning to drive in one day" (Non-Patent Document 2). [0003] As the inventor of "deep learning", Hinton declared in an interview with Axios in September 2017: "My point is to throw it all (backprop) and start over", which is Hinton's Boltz The dream of the Mann machine was shattered, and the black-box problem of "deep learning" was unsolvable, so it was not suitable f...

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 Applications(China)
IPC IPC(8): G06N3/04G06N3/08
CPCG06N3/08G06N3/047G06N3/045
Inventor 顾泽苍
Owner 顾泽苍
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