基于环形数据集的改进K-means聚类算法

Improved K-means Clustering Algorithm Based on Circular Data Set

K-means聚类算法是一种以距离为依据进行聚类的算法,原理简单且便于处理数据,但对于呈环形分布的数据集,传统的K-means算法的聚合结果并不理想。因此,本文提出了基于环形数据集的改进K-means聚类算法,引进了原点相对距离差的概念,将欧氏距离的计算改进为原点相对距离差的计算,使得数据集聚类为环形状。实验结果表明,改进的K-means聚类算法处理呈环形分布的数据集效果更好。

K-means clustering algorithm is a clustering algorithm based on distance.Its principle is simple and easy to process data.However,for data sets with ring distribution,the aggregation result of traditional K-means algorithm is not ideal.Therefore,this paper proposes an improved K-means clustering algorithm based on ring data set,and introduces the concept of relative distance diff erence of origin.The calculation of Euclidean distance is improved to the calculation of relative distance diff erence of origin,so that the data set is clustered into ring shape.The experimental results show that the improved K-means clustering algorithm is better in dealing with data sets with circular distribution.