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

Local density spectral clustering similarity measurement algorithm based on Self-tuning

A technology of similarity measurement and local density, applied in computing, computer components, instruments, etc., can solve problems such as enhancing weight values, and achieve the effect of eliminating the sensitivity of scale parameters and improving the ability to identify

Inactive Publication Date: 2015-01-28
INST OF ELECTRONICS & INFORMATION ENG IN
View PDF0 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Based on the ideas and existing problems of the above methods, we propose an improved spectral clustering method, that is, a local density spectral clustering similarity measurement algorithm based on Self-tuning, which mainly solves two problems: (1). Adaptive The calculation scale parameter σ of , does not need to manually set the value of the parameter; (2). Adaptively calculate the local density of the sample data, and enhance the weight value between the data points belonging to the same cluster

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 density spectral clustering similarity measurement algorithm based on Self-tuning
  • Local density spectral clustering similarity measurement algorithm based on Self-tuning
  • Local density spectral clustering similarity measurement algorithm based on Self-tuning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] A self-tuning-based local density spectrum clustering similarity measurement algorithm, the method is as follows:

[0029] (1) Assume that for an N-dimensional data set S={s 1 ,s 2 ,...,s M}∈R M×N , the number of samples is M, each sample s i is an N-dimensional data point, and its real number of clusters is C. The data set S is normalized, so that the feature data is normalized to [0, 1], and the influence of the order of magnitude between the data features is removed.

[0030] (2) Calculate the Euclidean distance between all data point pairs in the data set S, expressed as {d 1 , d 2 ,…,d n(n-1) / 2}.

[0031] (3) Calculate the value of the radius ε representing the local density according to the Euclidean distance d obtained in step (2), which satisfies that the average number of neighbors of the data point is 2%-3% of the total data number.

[0032] (4) According to the formula σ i =d(s i ,s k ) Calculate each data point s in the data set S i The local scal...

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 local density spectral clustering similarity measurement algorithm based on Self-tuning. Through analysis of a similarity measurement method, a local density measurement method based on data neighborhood is provided. The method can be adopted to adaptively measure the scale of data and deal with data set clustering problems with a complex structure. The method has a good clustering effect compared with traditional spectral clustering methods and Self-tuning methods.

Description

technical field [0001] The technical field of data clustering analysis of the present invention specifically relates to a self-tuning-based local density spectrum clustering similarity measurement algorithm, which can be used for cluster analysis of data, pictures and other information. Background technique [0002] Cluster analysis is an important unsupervised analysis method, and spectral clustering is a new clustering method. Based on the good clustering effect and perfect theoretical derivation of spectral clustering, it has been widely applied to data clustering. analyzing the problem. Traditional clustering algorithms, such as k-means algorithm and GMM algorithm, are suitable for convex spherical sample spaces. When the sample space is non-convex, the algorithm will fall into local optimum, but the spectral clustering method can converge well to the global Optimal, and without making any assumptions about the original structure of the data, spectral clustering can sho...

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): G06K9/62
CPCG06F18/2323
Inventor 陈雷霆蔡洪斌邱航关亚勇曹跃崔金钟卢光辉
Owner INST OF ELECTRONICS & INFORMATION ENG IN
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