Clustering method and device and terminal device

Abstract

The embodiment of the invention discloses a clustering method and device and a terminal device. The clustering method comprises the steps that firstly, when the number of neighbor objects of an object to be accessed in at least one neighborhood of multiple neighborhoods is not smaller than a corresponding preset threshold value, the object to be accessed is determined as a core object; secondly, the object to be accessed is classified into one type; thirdly, extension clustering is conducted on a directly density-reachable object of the object to be accessed in the assigned neighborhood; if the directly density-reachable object is not the core object, extension is stopped; fourthly, a next object to be accessed continues to be determined until clustering of all the objects to be accessed is completed. Due to the fact that multiple-neighborhood type judgment is used for judging whether the object to be accessed is the core object in the clustering method, it means that the limitation of the scanning radius and the minimum number of contained objects is loosened; in this way, the sensitivity of a clustering result to the two parameters including the scanning radius and the minimum number of contained objects is reduced, and the accuracy rate of the clustering result is increased.

Description

technical field [0001] The present disclosure relates to the technical field of data processing, and in particular to a clustering method, device and terminal equipment. Background technique [0002] Clustering is the process of dividing a collection of physical or abstract objects into multiple classes composed of similar objects, that is, the process of classifying objects into different classes or clusters. Objects in the same class have great similarity, and different classes There is a great deal of dissimilarity between the objects. [0003] Clustering methods include many types, among which, the density-based clustering method is different from other clustering methods in that it is not based on various distances, but based on density, as long as the density of points in an area is greater than a certain threshold , add it to the clusters that are close to it. This can overcome the shortcoming that the distance-based clustering algorithm can only find "circle-like" ...

AI Technical Summary

Problems solved by technology

However, the clustering results of this clustering algorithm are very sensitive to the values ​​of the parameters Eps and MinPts, that is, the values ​​of Eps and MinPts are different, resulting in different clustering results, resulting in uncertainty of the clustering results

Images