Local learning regularization-based depth matrix decomposition method and image clustering method

A matrix decomposition and local learning technology, applied in the field of image processing, can solve the problem that non-negative matrices are difficult to meet the multi-attribute properties of data, and achieve the effect of efficient use

Active Publication Date: 2019-02-12
JIANGSU UNIV OF TECH
View PDF4 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of the above problems, the present invention provides a deep matrix decomposition method and image clustering method based on local learning regularization, which effectively solves the technical problem that the existing non-negative matrix cannot meet the hierarchical structure required by the multi-attribute nature of the data

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
  • Local learning regularization-based depth matrix decomposition method and image clustering method
  • Local learning regularization-based depth matrix decomposition method and image clustering method
  • Local learning regularization-based depth matrix decomposition method and image clustering method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation manners of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention, and those skilled in the art can obtain other accompanying drawings based on these drawings and obtain other implementations.

[0027] For deep semi-non-negative matrix factorization (Deep Semi-NMF), the original dataset matrix Y that does not limit the positive and negative internal elements is given ± Decompose into (m+1) factors, so that the product of these (m+1) factors is as close as possible to the original data set matrix Y ± approximately equal. Under the limitation of non-negative coefficient matrix, deep semi-non-negative matrix regularization keeps information unchanged as much as possible, an...

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 provides a local learning regularization-based depth matrix decomposition method and an image clustering method, wherein the depth matrix decomposition method comprises the steps of S10obtaining a data matrix Y according to a picture to be clustered; S20 constructing an objective function based on the data matrix Y S30 using an iterative weighting method according to the objective function C to output a basis matrix Ni and a coefficient matrix Mi to complete the decomposition of the data matrix Y. Compared with the traditional absolute non-negative matrix factorization, not onlythe deep semi-non-negative matrix factorization is used to classify the data accurately and efficiently, but also some attribute information is combined to minimize the prediction cost of each region.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a deep matrix decomposition method and an image clustering method based on local learning regularization. Background technique [0002] In many cases, in order to deal with data classification and clustering problems, we often face the problem of dealing with high-dimensional data. The conventional solution to this kind of problem is to reduce the high-dimensional data, that is, to establish a certain data representation method, and use the low-dimensional data to approximate the high-dimensional data. Commonly used data representation methods are mainly divided into two categories: one is a linear representation method, and the other is a nonlinear representation method. For the linear representation method, only the low-dimensional manifold where the sampling data is located is linear, which is easy to use, but when the original data is linearly decomposed, the potent...

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
CPCG06F18/2133G06F18/23213
Inventor 舒振球孙燕武陆翼范洪辉李仁璞张杰汤嘉立
Owner JIANGSU UNIV OF 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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap