Unlock instant, AI-driven research and patent intelligence for your innovation.

A graph clustering method based on similarity transmission

A similarity and graph clustering technology, which is applied in the fields of pattern recognition and statistical data analysis to improve the accuracy of clustering and avoid post-processing operations.

Inactive Publication Date: 2019-05-17
NORTHWESTERN POLYTECHNICAL UNIV
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the problems existing in the existing graph clustering methods, the present invention proposes a graph clustering method based on similarity transfer

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
  • A graph clustering method based on similarity transmission
  • A graph clustering method based on similarity transmission
  • A graph clustering method based on similarity transmission

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] Now in conjunction with embodiment, accompanying drawing, the present invention will be further described:

[0039] refer to figure 1 , the specific implementation steps of the present invention are as follows:

[0040] Step 1, according to Nie et al. in the literature "F.Nie, X.Wang, M.Jordan, and H.Huang. The Constrained Laplacian Rank Algorithm for Graph-Based Clustering.AAAIConference on Artificial Intelligence, 1969-1976, 2016." The method of constructing a square matrix W with dimension n as the initialization similarity graph.

[0041] (1a) Assume that the data set contains n data points, each of which is a d-dimensional column vector, and the jth data point is represented by the symbol x j express. define data point x i and x j The distance is

[0042]

[0043] where e ij is the data point x i and x j The distance of ||·|| 2 is the vector two-norm.

[0044] (1b) For data x i , reordering its distance from all other points from small to large, so t...

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 graph clustering method based on similarity transmission. According to the algorithm, the similarity transmission is carried out between adjacent data points, a manifold structure is mined, an optimal graph capable of reflecting the topological relation between the data points is learned, through Laplace rank constraint, the learned optimal graph has a clear class structure (each connected component corresponds to one class), the post-processing operation is avoided, and the clustering accuracy is improved.

Description

technical field [0001] The invention belongs to the field of machine learning, and is particularly aimed at the problem of graph clustering, to learn the similarity between data, and to combine data with high similarity into the same class. The invention can be applied to statistical data analysis, pattern recognition and other aspects. Background technique [0002] With the in-depth development of the big data era, data mining has gradually become a hot spot in the field of machine learning. In data mining technology, cluster analysis is one of the widely researched subjects. The so-called clustering is based on the similarity between data objects, distinguishing them, and merging data into different categories. Data classified into the same category have high similarity, while data in different categories have low similarity. The clustering method does not rely on prior information, so it is an unsupervised learning method. Because of its unsupervised nature, cluster a...

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): G06K9/62
Inventor 李学龙陈穆林王琦
Owner NORTHWESTERN POLYTECHNICAL UNIV