Spectral clustering method for automatically determining number of clusters based on neighboring point method

Abstract

A spectral clustering method for automatically determining the number of clusters based on a neighboring point method comprises the steps of 1) normalizing all dimensions of a data set; 2) calculating an interval sparse distance matrix by a neighboring point method and defining the matrix as local scale parameters of distance mean values of the neighboring points to obtain a whole sparse similarity matrix; 3) determining the local density of each data point and the minimum distance to other points with a higher local density by calling a CCFD method, and obtaining the number of singular points generated by the fitting outside a confidence interval; 4) calculating a degree matrix D and a Laplacian matrix L according to a formula and extracting an eigenvector group by eigen decomposition of L; 5) outputting clustering results; and 6) selecting and outputting the clustering result with the optimal number of neighboring points corresponding to the maximum Fitness function value. According to the invention, the local scale parameter of each data point can be estimated according to data distribution, the number of clustering centers is automatically determined, and the parameter adaptation of the number of neighboring points is realized.

Description

technical field [0001] The invention belongs to the field of spectral clustering algorithms, in particular to a spectral clustering method for automatically determining the number of clusters based on an adjacent point method. Background technique [0002] Clustering is to gather objects with higher similarity among physical or abstract objects in the same class, and assign objects with lower similarity to different classes, so that the relationship between all objects in the cluster formed by the same cluster have high similarity, while the similarity between objects in different clusters is low. Clustering analysis technology has a very wide range of applications in the fields of gene expression analysis and image processing. Clustering algorithm is one of the main methods to deal with image segmentation, and has been widely used in image segmentation. In the past few decades, spectral clustering algorithms have also shown obvious advantages in image segmentation and dat...

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the high spatial complexity of existing spectral clustering algorithms, it is difficult to select appropriate scale parameters to well reflect the spatial distribution characteristics of complex data sets, and it is necessary to manually input the number of clusters. Aiming at these problems, the present invention A spectral clustering method based on the adjacent point method is proposed, which automatically selects the number of clusters and estimates the local scale parameters of each data point according to the data distribution, which effectively reduces the space complexity and selects appropriate scale parameters to It can well reflect the spatial distribution characteristics of complex data sets, and does not need to manually input the number of clusters

Images