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

Molecular structure diagram retrieval method based on evolutionary computation multi-view fusion

A molecular structure map and evolutionary computing technology, applied in still image data retrieval, computing, computer components and other directions, can solve the problem that the depth model cannot capture the features well

Active Publication Date: 2020-11-06
SHANXI UNIV
View PDF8 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Effective feature representation plays a core and fundamental role in molecular structure retrieval performance, however, a single deep model cannot capture its features well

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
  • Molecular structure diagram retrieval method based on evolutionary computation multi-view fusion
  • Molecular structure diagram retrieval method based on evolutionary computation multi-view fusion
  • Molecular structure diagram retrieval method based on evolutionary computation multi-view fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] Such as Figure 1 to Figure 4 As shown, a molecular structure graph retrieval method based on evolutionary computing multi-view fusion, including the following steps:

[0044] Step 1, data enhancement: Given a molecular structure graph data set, expressed as D={(x i ,y i )|1≤i≤|D|}, where x i represents the molecular structure diagram, y i means x i categories; first, use data augmentation methods to enhance the molecular structure graph dataset D; then, uniformly scale the graphs in the enhanced dataset to w×h size, where: w represents the image width, h represents the image height; finally, obtain the enhanced The copy of each picture in the data set with different reduction ratios is enlarged to w×h by padding 255 pixels in the outer margin. The data set composed of all molecular structure diagrams generated through this process is denoted as D * ;

[0045] Step 2, train the multi-view feature extractor: Given a set of deep convolutional networks N={AlexNet, VG...

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 molecular structure diagram retrieval method based on evolutionary computation multi-view fusion. The method comprises the following steps: 1, enhancing data by using operations such as rotation, translation, scaling and the like; 2, training a plurality of depth models by using the enhanced data to serve as a multi-view feature extractor; 3, extracting multi-view features of the enhanced data by using the trained multi-view extractor; 4, searching a multi-view fusion model with excellent performance through an evolutionary algorithm to obtain fusion features of themolecular structure diagram classification model and the data set; and step 5, directly using the classification model or using the fused features to calculate the obtained sorting value of the similarity between the to-be-retrieved graph and the molecular graph in the retrieval library to complete retrieval of the to-be-retrieved molecular structure graph. The invention is used for solving the problem of direct image-based chemical molecular structural formula retrieval without depending on molecular structural formula coded characters in the field of chemical informatics.

Description

technical field [0001] The invention relates to the field of chemical molecular structure retrieval of chemical informatics, in particular to a molecular structure diagram retrieval method based on evolutionary calculation and multi-view fusion. Background technique [0002] Chemical structural formula retrieval is one of the core tasks in the field of chemical informatics, and it is a chemical information search method that uses the input chemical molecular structure graphics as the retrieval content. This method is currently a common search method for chemical and chemical talents when conducting scientific research or ordering chemical reagents. [0003] At present, chemical structure retrieval mainly adopts the method of pre-defined molecular structure coding. For example, based on the structural formula retrieval based on Simplified Molecular Linear Input Specification (SMILES), this method first needs to split the chemical structure into fragments represented by symbo...

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): G06F16/55G06F16/583G06K9/42G06K9/62G06N3/04G06N3/08
CPCG06F16/583G06F16/55G06N3/08G06V10/32G06N3/045G06F18/25G06F18/241
Inventor 梁新彦郭倩钱宇华朱哲清彭甫镕
Owner SHANXI UNIV
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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com