Clustering analysis method based on sample density and self-adaptive adjustment of clustering center

Abstract

The invention relates to a clustering analysis method based on sample density and self-adaptive adjustment of a clustering center. Firstly, a clustering number K and the clustering center are initialized based on the sample density and the sample neighborhood, then the clustering number K is adaptively adjusted by adopting a nearest neighbor cluster merging method, and finally, the optimal clustering result is obtained, so that the effectiveness of the clustering analysis result is greatly improved.

Description

technical field [0001] The invention relates to the field of unsupervised learning cluster analysis in machine learning, in particular to a cluster analysis method based on sample density and self-adaptive adjustment of cluster centers. Background technique [0002] At present, the most widely researched and applied is the clustering method based on partitioning. The K-means algorithm is a typical clustering algorithm based on the idea of ​​partitioning and a clustering algorithm based on distance. The characteristics of fast speed and strong local search ability have been studied by many experts and scholars at home and abroad for many years, and have been widely used in many industrial and commercial fields. However, the K-means clustering algorithm needs to manually determine the K value of the clustering category number in advance, and the clustering result is greatly affected by the initial value, which limits its role in solving practical problems to a certain extent. ...

AI Technical Summary

Problems solved by technology

However, the K-means clustering algorithm needs to manually determine the K value of the clustering category number in advance, and the clustering result is greatly affected by the initial value, which limits its role in solving practical problems to a certain extent.

[0003] When using the K-means algorithm to process data sets, the number of clusters is often not known in advance, and the estimated number of clusters needs to be obtained based on prior knowledge or analysis of the data set. If the data set is very large or the data structure is very complex When , it is very expensive to analyze the data and it is difficult to determine the optimal number of clusters

In addition, due to the random selection of the initial clustering center of the algorithm, the clustering result is unstable, and it is easy to fall into a local optimal solution, resulting in a large deviation between the clustering effect and the actual feature distribution of the data

Images