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

Supervised local projection method for mining similarity between data

A local projection, supervised technology, applied in the fields of electrical digital data processing, natural language data processing, special data processing applications, etc.

Inactive Publication Date: 2020-08-11
GUANGXI NORMAL UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In addition, in terms of the quality of the obtained similarity matrix, although these three algorithms perform well on the data sets described in their attached papers, the quality of the similarity matrices obtained on some UCI data sets is not very good. Ideal, the quality needs to be improved, and more new methods are needed to make up for it

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
  • Supervised local projection method for mining similarity between data
  • Supervised local projection method for mining similarity between data
  • Supervised local projection method for mining similarity between data

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0070] A supervised local projection method for mining similarity between data, including the following steps:

[0071] 1) Obtain data and propose a mathematical model: obtain sample data X∈R n×d , label data Y∈R n×d and the maximum number of neighbors k for each sample, where X∈R n×d Is a matrix with n samples, each sample has d features; Y∈R n×c Is a one-hot matrix, indicating that there are n samples of labels, and the number of labels is c;

[0072] Then the mathematical model (1) is proposed:

[0073]

[0074] where W∈R d×c is a projection matrix of size d×c; S∈R n×n is the similarity matrix, S i ∈R 1×n Indicates the data of the i-th row in S, and the elements of the initialization similarity matrix S are all 0; λ 1 is a constant coefficient, initialized to λ 1 =1; I is the identity matrix; L=D-S∈R n×n is the Laplacian matrix of the similarity matrix S, D∈R n×n is the degree matrix of the similarity matrix S, rank(L) means to rank L, ||S|| F Indicates the c...

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 supervised local projection method for mining similarity between data. The method comprises the following steps: 1) acquiring data and proposing a mathematical model; 2) initializing, 3) updating, 4) converting, 5) calculating the maximum characteristic value alpha of an intermediate matrix variable, 6) calculating the value of loss1, 7) simplifying a mathematical model,8) projecting, 9) calculating the value of loss2, and 10) calculating the loss value of the mathematical model. The method can improve the classification accuracy and reduce the calculation complexityof batch operation.

Description

technical field [0001] The present invention relates to the field of deep learning, machine learning or artificial intelligence, in particular to a supervised local projection method for mining the similarity between data. Background technique [0002] Although there is no completely independent method to mine the similarity between data, there are related sub-modules in the clustering field of machine learning to accomplish this purpose. For example, Professor Nie Feiping of Northwestern Polytechnical University's "Optical Image Analysis and Learning Center" published papers [1] in 2014 and papers [2] in 2016, both of which contain methods for mining the similarity between data. Abbreviated as 2014CAN and 2016CLR. In addition, in the field of machine learning, the most common and easiest way to obtain the similarity between data is the k-neighbor algorithm. The k-neighbor algorithm is also called the Knn algorithm [3]. Its idea is to calculate the Euclidean distance betwe...

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): G06F16/35G06F17/16G06F40/194G06K9/62G06N3/04
CPCG06F16/35G06F17/16G06F40/194G06N3/045G06F18/22
Inventor 朱晓峰张北贤陈林君詹猛猛张乐园张师超
Owner GUANGXI NORMAL 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