An improved DPC clustering algorithm and system based on a symmetric neighbor relation

Abstract

The invention discloses an improved DPC clustering algorithm and system based on a symmetric neighbor relation, and the algorithm comprises the following steps of S1, obtaining the influence space ofeach sample point; S2, calculating the product of the sample point density and the distance; and S3, performing clustering operation according to the width-first traversal mode. According to the method, parameters can be reasonably selected, the Miller effect is avoided, the clustering effect is improved, and particularly, a very good clustering effect can be obtained on a data set with non-uniform density.

Description

technical field [0001] The invention relates to the field of clustering algorithms, in particular to an improved DPC clustering algorithm and system based on symmetric neighbor relationship. Background technique [0002] Clustering is an important data mining method, which belongs to a description class of data mining methods. Clustering divides data objects into multiple groups or clusters, objects in the same group are similar to each other, and objects in different groups are not similar to each other. Unlike classification, clustering is an unsupervised learning method that requires no prior knowledge and does not require labeling of the data. There are various clustering algorithms, including partitional clustering, hierarchical clustering, density-based clustering, distribution-based clustering, and spectral clustering. The DPC algorithm is a density-based clustering algorithm that was published in the journal Science in 2014. The algorithm finds the center of the c...

AI Technical Summary

Problems solved by technology

But at the same time, it still has some defects, such as the parameter truncation distance d c It is difficult to determine the value of , which has a serious impact on the clustering results; if the distribution of non-central points is unreasonable, it is easy to produce the "Domiller" effect; the clustering results shown in uneven density data sets are not ideal , especially the manifold dataset

[0003] For example, the invention patent application with the application publication number CN107392249A discloses a K-nearest neighbor similarity-optimized density peak clustering method, which uses K-nearest neighbors to detect pointing points when assigning each point, and finally assigns the remaining points to pointing points. The family of points to improve the accuracy and scope of application of the DPC algorithm, but after sorting the distances between all points from small to large, take the 2% to 5% small distance as the cut-off distance d c , there is still a truncation distance d c It is difficult to determine the value of

Images