Supercharge Your Innovation With Domain-Expert AI Agents!

Minimum coverage set problem solving method based on graph deep learning

A problem-solving and deep learning technology, applied in neural learning methods, neural architectures, biological neural network models, etc., can solve problems such as difficulty in guaranteeing algorithm errors

Pending Publication Date: 2021-09-07
SHANGHAI UNIVERSITY OF ELECTRIC POWER
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This is a classic NP-Hard problem. In theory, there is no optimal solution in polynomial time. The usual solution is to design a heuristic algorithm, which needs to be based on specific examples and knowledge in the field, and the error of the algorithm is in Difficult to guarantee on larger scale graphs

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
  • Minimum coverage set problem solving method based on graph deep learning
  • Minimum coverage set problem solving method based on graph deep learning
  • Minimum coverage set problem solving method based on graph deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0043] The present invention aims to solve a classic NP-hard graph combinatorial optimization problem minimum cover set, and its variants have a wide range of applications in reality, such as recommendation systems, viral promotion on social networks, medical services and so on. In solving the MVC problem, the present invention will focus on designing a popular model of approximate heuristic algorithm, called the greedy algorithm. By adding vertices step by step, a partial solution S is constructed. This process is called the greedy algorithm on the graph. For the MVC problem, according to the input graph, the vertex with the highest value is added first. Inspired by reinforcement learning, we learn to select the optimal vertex through continuous exploration and feedback.

[0044] refer to Figure 1~4 , as an embodiment of the present invention, provides a method for solving the minimum cover set problem based on graph deep learning, including:

[0045] S1: Use the networkx m...

Embodiment 2

[0083] refer to Figure 5-6 It is another embodiment of the present invention. In order to verify and illustrate the technical effect adopted in this method, this embodiment adopts a traditional technical solution to carry out a comparative test with the method of the present invention, and compares the test results by means of scientific demonstration to verify that this method has real effect.

[0084] In order to test the effect of the present invention on solving the MVC problem, the following experiments were designed. To evaluate the solution quality of the generated instances, the approximate ratio between the inventive method and the classical heuristic is used as a reference. The calculation method is as follows:

[0085]

[0086] Wherein C (S, G) is the solution value of the present invention, Heuristic (G) is the solution value of the classic heuristic method, and the lower the approximate ratio is, the better the effect is. The experiment of the present inven...

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 a minimum coverage set problem solving method based on graph deep learning, and the method comprises the steps: generating an input graph through employing a network workx module in python, and initializing the vertex features of the input graph through combining the state of a vertex and in-degree and out-degree information; according to the input graph generation rule, sampling neighbor vertexes of each vertex in the input graph to embed structural information of the graph; after neighbor vertex sampling, transmitting characteristics of neighbor vertexes to a current node, splicing vertex and performing L2 regularization on the vertex vectors after splicing operation; and setting reinforcement learning parameters, carrying out reinforcement learning optimization based on a regularization processing result, and solving a minimum coverage set. The model has good expansibility and can be used for large-scale real scenes.

Description

technical field [0001] The present invention relates to Minimum Vertex Cover (MVC), the technical field of graph deep learning, and in particular to a method for solving a minimum cover set problem based on graph deep learning. Background technique [0002] MVC problems can be classified as combinatorial optimization problems on graphs. In the field of applied mathematics and theoretical computers, graph combinatorial optimization refers to finding an optimal subset according to a certain goal in a finite set with certain characteristics. Mathematical programming methods. MVC can be defined as given an undirected graph G(V, E), find the smallest subset of vertices S ∈ V such that every edge in the graph is incident to at least one vertex in the set S. This is a classic NP-Hard problem. In theory, there is no optimal solution in polynomial time. The usual solution is to design a heuristic algorithm, which needs to be based on specific examples and knowledge in the field, and...

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): G06N3/04G06N3/08
CPCG06N3/08G06N3/044G06N3/045
Inventor 杜海舟严宗
Owner SHANGHAI UNIVERSITY OF ELECTRIC POWER
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