Supercharge Your Innovation With Domain-Expert AI Agents!

Class tree-based deep learning method and neural network system

A deep learning and neural network technology, applied in neural learning methods, biological neural network models, etc., can solve problems such as low accuracy, improve accuracy, reduce structural complexity, training and application difficulty, and reduce the number of input items and the effect of outputting the number of categories

Active Publication Date: 2020-09-01
SOUTH CHINA NORMAL UNIVERSITY
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Based on this, it is necessary to provide a deep learning method and neural network system based on category trees to solve the problem of low accuracy when the number of input items and output categories of traditional deep learning neural networks are too large.

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
  • Class tree-based deep learning method and neural network system
  • Class tree-based deep learning method and neural network system
  • Class tree-based deep learning method and neural network system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] The technical solution of the present invention will be described below in conjunction with the accompanying drawings.

[0020] Such as figure 1 As shown, the present invention provides a kind of deep learning method based on class tree, may comprise the following steps:

[0021] S1, obtaining the category tree corresponding to the output data in the training data;

[0022] Wherein, the manner of obtaining the category tree may be automatic construction.

[0023] In one embodiment, a class tree such as figure 2 shown. Output data includes, but is not limited to, output labels. When the output data in the training data includes the output labels "pig", "dog", "grass", "tree", "algae", "man", "woman", these output data correspond to figure 2 The leaf node of the category tree in the middle. The output label "creature" in the training data corresponding to the biological category in the category tree; the output label in the training data corresponding to the anima...

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 category tree-based deep learning method and neural network system. Through multi-level deep learning neural network layer-by-layer training and layer-by-layer application, the number of input items and the number of output categories are reduced, thereby reducing the number of deep learning neural networks at all levels. The structural complexity of the network and the difficulty of training and application reduce the computational complexity of the deep learning neural network and improve the accuracy of the deep learning neural network.

Description

technical field [0001] The invention relates to the technical field of deep learning neural networks, in particular to a category tree-based deep learning method and a neural network system. Background technique [0002] A deep learning neural network is trained by input data and output data. When the number of input items and output categories of the traditional deep learning neural network are too large, the deep learning neural network will be too complex, and the amount of input data and output data will be too large, which will make the structure of the deep learning neural network too complex and deep. The amount of calculation in the training of the learning neural network is too large, and it even fails to converge, which affects the quality of the training, which in turn leads to a decrease in the accuracy of the deep learning neural network during application. [0003] To sum up, the accuracy of traditional deep learning neural network is low when the number of 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): G06N3/08
CPCG06N3/08
Inventor 朱定局
Owner SOUTH CHINA NORMAL UNIVERSITY
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