Self-adaptive K-means clustering method based on local density and ball hash

Abstract

The invention discloses a self-adaptive K-means clustering method based on local density and ball hash. Firstly, a target data set is normalized, the local density and median of samples in the data set are calculated to determine the clustering category number k, then k optimized initial clustering centers are obtained according to the quartile and the ball hash value under the local density of the adjacent samples, and finally clustering of the data set is completed through a K-means clustering algorithm. According to the method, the category number k and the initial clustering center of thedata set are automatically determined, the defects that the category number needs to be set in advance and the initial clustering center needs to be set at will in a traditional K-means clustering algorithm are overcome, and K-means clustering of the data set can be carried out in a self-adaptive mode.

Description

technical field [0001] The invention relates to the field of data mining, in particular to an adaptive K-means clustering method based on local density and spherical hashing. Background technique [0002] In the definition of the field of data mining, a data set is a collection of data, clustering is a process of classifying and organizing data members that are similar in some aspects in a data set, and clustering is a process of discovering such data. Intrinsically structured techniques, clustering techniques are often referred to as unsupervised learning. [0003] Cluster analysis is an effective data mining method, which can express the internal structure characteristics of data, and the scatter diagram generated by the clustering results is also an effective means of data visualization. K-means clustering algorithm (K-means clustering algorithm) is a cluster analysis algorithm for iterative solution. Its steps are to randomly select k objects as initial cluster centers,...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above-mentioned deficiencies in the prior art, an adaptive K-means clustering method based on local density and spherical hashing provided by the present invention solves the optimization of the number of categories k and the initial cluster center existing in the K-means clustering process The selection problem, this method can adaptively select the number of categories k and the initial cluster center for different data sets, and improve the adaptability and clustering effect of K-means clustering

Images