Network structure searching method, device and apparatus and storage medium

A technology of network structure and search method, which is applied in the field of devices, equipment and storage media, and search method of network structure, which can solve the problems of long search time and achieve the effect of shortening search time and training time

Pending Publication Date: 2021-09-17
SF TECH
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing neural network architecture search algorithm has the problem of long search time

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
  • Network structure searching method, device and apparatus and storage medium
  • Network structure searching method, device and apparatus and storage medium
  • Network structure searching method, device and apparatus and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0138] An implementation manner of determining the target network performance is: taking the average value of the first network performance and the second network performance as the target network performance of the first network structure. Specifically, the target network performance of the first network structure is calculated by the following formula (4), wherein the formula is:

[0139]

[0140] Among them, P represents the target network performance of the first network structure, and P e Indicates the network performance (that is, the first network performance) predicted by the first network structure through the performance predictor, P l Indicates the network performance (that is, the second network performance) detected on the verification set after the first network structure is trained by the training data set.

[0141] Another implementation manner of determining the target network performance is as follows: firstly, respectively determine the weight coefficien...

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 embodiment of the invention discloses a network structure searching method, device and apparatus and a storage medium. The network structure searching method comprises the following steps: acquiring a first network structure according to a preset searching strategy; determining target network performance of the first network structure through a preset performance prediction model; updating the searching strategy according to the target network performance, obtaining a second network structure according to the updated searching strategy until meeting preset search stopping conditions. According to the embodiment of the invention, the training time of network structures is shortened, and the search time of network structures is reduced.

Description

technical field [0001] The embodiment of the present application relates to the technical field of machine learning, and specifically relates to a network structure search method, device, device, and storage medium. Background technique [0002] With the rapid development of deep learning technology, deep learning technology has been widely used, such as in image recognition, speech recognition and so on. [0003] However, it is very difficult to manually design a neural network with excellent performance. For example, the structural design, training strategy, and parameter adjustment methods of the neural network often require researchers to spend a lot of manpower and material resources on experimental exploration. [0004] At present, neural network architecture search algorithms are proposed in the prior art, aiming at automatically completing the search of neural network architectures, so as to solve the problems existing in artificially designed neural networks. Howev...

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): G06N3/04G06N3/08G06K9/62G06N20/00
CPCG06N3/084G06N20/00G06N3/044G06N3/045G06F18/24
Inventor 杨杰
Owner SF TECH
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