Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Automatic classification method of convolutional neural network constructed based on incremental branch growth

A convolutional neural network and automatic classification technology, applied in the field of neural networks, can solve the problems of tedious parameter adjustment process and reduce task efficiency, so as to avoid the parameter adjustment process and speed up the search process.

Pending Publication Date: 2020-04-07
GUILIN UNIVERSITY OF TECHNOLOGY
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The task of using convolutional neural networks for automatic classification is one of the applications of convolutional neural networks. In practical applications, convolutional neural networks need to be learned. By manually adjusting the parameters of the neural network, the target features in the same classification can be recognized. , but such a process requires a cumbersome parameter tuning process, which greatly reduces the efficiency of the task

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
  • Automatic classification method of convolutional neural network constructed based on incremental branch growth
  • Automatic classification method of convolutional neural network constructed based on incremental branch growth

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] The present embodiment provides an automatic classification method of a convolutional neural network based on incremental branch growth, such as figure 1 , including the following steps:

[0035] S1: Construct and initialize the convolutional neural network model G m ;

[0036] S2: Convolutional neural network model G m Perform training to obtain the trained model G′ m ;

[0037] S3: Convolutional neural network model G m Carry out next-generation network growth to obtain the next-generation convolutional neural network model G m+1 ;

[0038] S4: Convolutional neural network model G m+1 Perform training to obtain the trained model G′ m+1 ;

[0039] S5: If the model G′ m and model G′ m+1 The difference in classification test accuracy is less than the preset threshold, then use the model G′ m+1 Complete the classification task; if the model G′ m and model G′ m+1 The difference between the classification test accuracy is not less than the preset threshold, th...

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 an automatic classification method of a convolutional neural network constructed based on incremental branch growth. The method comprises the steps of S1, constructing and initializing a convolutional neural network model Gm; s2, training the convolutional neural network model Gm to obtain a trained model G'm; s3, performing next-generation network growth on the convolutional neural network model Gm to obtain a next-generation convolutional neural network model Gm + 1; s4, training the convolutional neural network model Gm + 1 to obtain a trained model G'm + 1; s5, if the difference between the classification test precision of the model G'm and the classification test precision of the model G'm + 1 is smaller than a preset threshold value, using the model G'm + 1 tocomplete a classification task; and if the difference between the classification test precision of the model G'm and the classification test precision of the model G'm + 1 is not smaller than a preset threshold value, m <-m + 1, and returning to the step S3. According to the method, a tedious parameter adjustment process in a neural network construction process can be avoided, and a deep convolutional neural network model suitable for a specific classification task can be more efficiently and automatically constructed.

Description

technical field [0001] The present invention relates to the field of neural networks, and more specifically, to an automatic classification method for convolutional neural networks constructed based on incremental branch growth. Background technique [0002] With the development of artificial intelligence and computer technology, the application of convolutional neural network in reality is more and more common. The task of using convolutional neural networks for automatic classification is one of the applications of convolutional neural networks. In practical applications, convolutional neural networks need to be learned. By manually adjusting the parameters of the neural network, the target features in the same classification can be recognized. , but such a process requires a cumbersome parameter tuning process, which greatly reduces the efficiency of the task. Contents of the invention [0003] The present invention provides an automatic classification method of a conv...

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): G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F18/24
Inventor 黄琳杨铁军
Owner GUILIN UNIVERSITY OF TECHNOLOGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products