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

Multi-view low-rank sparse subspace clustering method based on exclusive regularization

A clustering method and exclusive technology, applied in the field of data processing, can solve the problem of low multi-view clustering performance, and achieve the effect of ensuring low rank and sparsity

Active Publication Date: 2020-04-21
GUANGDONG UNIV OF PETROCHEMICAL TECH
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in the above existing technologies, the multi-view clustering performance is still low, and the spatial clustering method based on exclusive regularization still has room for optimization and improvement.

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
  • Multi-view low-rank sparse subspace clustering method based on exclusive regularization
  • Multi-view low-rank sparse subspace clustering method based on exclusive regularization
  • Multi-view low-rank sparse subspace clustering method based on exclusive regularization

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0017] refer to figure 1 , the specific implementation process of the present invention is as follows:

[0018] Step 1. For the given multi-view data X, according to the low-rank sparse subspace clustering method, low-rank and sparse constraints are respectively performed on the representation coefficient matrix Z of each view data. The model is

[0019]

[0020] s.t.X (v) =X (v) Z (v) +E (v) ,diag(Z (v) )=0,

[0021] where Z (v) ∈ R N×N ,E (v) ∈ R D×N Indicates the noise of the data, λ 1 ,λ 2 are non-negative low-rank, sparse regularization parameters, respectively.

[0022] Step 2, for different view data, pairwise exclusive constraints are imposed on the representation coefficient matrix Z respectively. Its exclusive constraints are expressed as where is the Hadamard product. due to l 0 The norm is non-convex, so use its convexity instead to change the above constraint to The purpose of this constraint is to make the data in different views as diverse ...

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 multi-view low-rank sparse subspace clustering method based on exclusive regularization. Low rank and sparsity constraints of a coefficient matrix are represented through subspaces, and low rank and sparsity of the coefficient matrix are ensured at the same time. Besides, exclusive constraint is carried out on the subspace coefficient matrixes of different views, so thatthe views are as diversified as possible, and information among different views is fully utilized. Then, an incidence matrix shared by all views is constructed by utilizing the obtained coefficient matrix; and finally, clustering operation is performed on the incidence matrix by a spectral clustering method. According to the invention, an alternating direction multiplier method is adopted to solvethe model. Compared with the experiment of the existing multi-view subspace clustering method on different data sets, the method provided by the invention has better clustering performance.

Description

technical field [0001] The invention belongs to the field of data processing, in particular to a multi-view low-rank sparse subspace clustering method based on exclusive regularization. Background technique [0002] In the past ten years, subspace clustering has become a hot research topic. In computer vision and image processing applications, data are often high-dimensional, which can often be represented by low-dimensional subspaces. When the data points are all distributed in a single subspace, its main purpose is to find the basis of the subspace and a low-dimensional representation of the data points. Commonly used methods are principal component analysis, independent component analysis, and nonnegative matrix factorization. On the other hand, when the data points are distributed in different subspaces, it can be solved by standard clustering algorithms, such as k-means clustering algorithm. However, these methods are based on the premise that the data points are dis...

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
CPCG06F18/2321
Inventor 曹飞龙张清华
Owner GUANGDONG UNIV OF PETROCHEMICAL TECH
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